FINK: BASAL EUTELEOSTS 



205 



coregonins. In a number of features, including the morphology 

 of the nares, Prosopium stands as the sister group of other cor- 

 egonins, and this, plus the presence in the outgroup Salmoninae 

 and Thymallus of exoccipital participation in the condyle, im- 

 plies that phylogenetically derived coregonins have secondarily 

 lost that morphology. As noted by Fink and Weitzman (1982), 

 the condyle structure as found in salmonids is found also in 

 neoteleosts. It is also present in Lepidogalaxias (see below) and 

 in some osteoglossomorphs. I do not wish to belabor the possible 

 importance of this character, especially since more careful on- 

 togenetic and morphological studies need to be done and other 

 characteristics evaluated. 



A few observations from my survey of salmonids may be 

 added here. I have found but two characters in the literature 

 which diagnose the coregonins; one of these needs modification 

 and the other needs to be more concisely put. Lack of maxillary 

 teeth has been used to diagnose the group, relative to other 

 salmonids (Norden, 1 96 1 ), but this needs to be emended to lack 

 of the teeth in adults, since I have found maxillary teeth in 

 Prosopium of around 19 mm SL. I have not yet examined spec- 

 imens this small of other coregonins so do not know the gen- 

 erality of this primitive state. The other character is reduction 

 in the teeth in general; this needs to be quantified relative to 

 the outgroups. 



The salmonins and Thymallus can be placed together based 

 on lack of ossification of the supraethmoid (hypethmoid of Nor- 

 den, 1961; Behnke, 1968), and apparently on yolk character- 

 istics, and larval size (Kendall and Behnke, this volume). Re- 

 garding other relationships within salmonids, I have nothing to 

 add. 



Lepidogalaxias.— The position of Lepidogalaxias is controver- 

 sial. I remain unconvinced by Rosen's (1974) hypothesis that 

 the genus belongs with the esocoids. When I previously dis- 

 cussed this genus (Fink and Weitzman, 1982), I had not seen 

 any specimens, but R. M. McDowall has generously made sev- 

 eral available for dissection and clearing and staining. There is 

 no question that this little fish is a potpourii of contradictory 

 and reductive characters and it is no wonder that it has been so 

 difficult to place. Pursuing the potential of relationship of this 

 species to galaxiids, extensive comparisons with members of 

 that group have been made. Lepidogalaxias shares a host of 

 reductive characters with galaxiids. NVhile these may indeed be 

 synapomorphous traits, in cases where extensive paedomor- 

 phosis is suspected, and this appears to be so in the morpho- 

 logical similarities involved, one hopes to find some innovative, 

 non-reductive characters which supply evidence for grouping. 

 I have found two such characters which suggest that Lepido- 

 galaxias is related to neither esocoids nor osmeroids, but rather 

 may be the sister group of the Neoteleostei, as diagnosed by 

 Rosen (1973) and Fink and Weitzman ( 1 982). This is supported 

 by the presence in Lepidogalaxias of two non-reductive traits, 

 a retractor dorsalis muscle and occipital condyle composed of 

 both the basioccipital and exoccipital bones. As discussed just 

 above and by Fink and Weitzman ( 1 982), the latter trait is also 

 shared with salmonids. Lepidogalaxias lacks a rostral cartilage 

 or its homologue and type 4 teeth (hinged teeth with a posterior 

 axis of rotation. Fink, 1981) and this would prevent its place- 

 ment within the neoteleostean assemblage. Placing Lepidoga- 

 laxias as the neoteleostean sister group and leaving salmonids 

 as their sister taxon presumes either that rostral cartilage homo- 

 logues in the salmonids have been lost in Lepidogalaxias or are 



EUTELEOSTEI 



NEOTELEOSTEI 



EURYPTERYGII 



OSTEOGLOSSOMORPHA 

 ELOPOMORPHA 

 CLUPEOMORPHA 

 ESOCOIDEI 

 OSTARIOPHYSI 

 ARGENTINOIDEI 

 OSMEROIDEI 

 SALMONIDAE 



LEPIDOGALAXIAS 



STOMIIFORMES 



AULOPIFORMES 



MYCTOPHIFORMES 



ACANTHOMORPHA 



Fig. 108. Summary cladogram of relationships and characters dis- 

 cussed in the text. 



not homologues after all. This ambiguity is reflected in Fig. 108 

 by a trichotomy. Clearly, more work remains to be done before 

 we can be really confident in the phylogenetic placement of this 

 intriguing fish. 



Lepidogalaxias can be diagnosed by a number of characters, 

 the most striking of which is fusion of the frontal bones into a 

 single ossification (Rosen, 1974, Fig. 40B). In their comments 

 on this species. Fink and Weitzman ( 1 982) noted that there was 

 a disagreement about whether there are mesopterygoid teeth 

 present; Rosen's statement that teeth are lacking is cortect. 



Stomiiformes. — Vmk and Weitzman (1982) recently examined 

 the monophyly and relationships of stomiiforms to the other 

 basal euteleosts and corroborated Rosen's (1973) hypothesis 

 that they are the sister group to the rest of the Neoteleostei, 

 removing them from the "salmoniforms." This placement is 

 supported by several apomorphic traits, including presence of 

 retractor dorsalis muscles and type 4 tooth attachment, as well 

 as exoccipital participation in the cranial condyle and a rostral 

 cartilage. Weitzman (1974) presented a hypothesis of relation- 

 ships at the "family" level within the stomiiforms, as well as a 

 detailed phylogeny of the Stemoptychidae. In this volume, I 

 present a generic-level phylogeny for the barbeled stomiiforms 

 (Family Stomiidae) and some brief comments on the "gonosto- 

 matid-photichthyid" genera. Weitzman is currently working on 

 relationships of the latter fishes and has made considerable com- 

 ments in this volume (see Ahlstrom, Richards and Weitzman, 

 this volume). 



