ESPECIALLY MYCTOPHOIDS 221 
first pelvic ray, but this structure is absent in Aulopus. The branchiostegal ray 
number is variable. There are 8 or g rays in Sardimioides, but Aulopus has approxi- 
mately 15. The vertebral column is a variable feature in that the elements are 
incompletely fused in Sardinioides but are much more closely united in Auwlopus. 
Although the total vertebral count is different (30 in Sardinioides, 50 in Aulopus), 
the number of caudal vertebrae is very similar, with 16 in Sardinioides and 14 in 
Aulopus. Both genera have feebly ctenoid scales, but in Sardinioides cycloid scales 
are also present. Although the basic composition of the caudal skeleton is the same 
in the two genera, that of Sardinioides (Text-fig. 72) has a fairly wide gap between 
the second and third hypurals, whereas in Aulopus (Text-fig. 93) hypurals two and 
three contact each other. 
These differences are only minor ones and the two genera are extremely closely 
related although being considerably separated in time. It is difficult to derive 
Aulopus directly from Sardinioides because of the possible absence of a supraorbital, 
the small number of branchiostegal rays and the presence of a gap between the 
second and third hypurals in Sardinioides. The hypural gap is also seen in the 
Ctenothrissiformes (Text-fig. 77), in certain polymixioid beryciforms (Patterson, 
1968b), and in the Paracanthopterygii (sensu Greenwood, ef al., 1966). No obvious 
functional significance can be attached to this feature, the fin outline remaining 
the same irrespective of whether the gap is present or absent. Many of the more 
advanced myctophiforms exhibit a slight gap but never to the extent shown in 
Sardinioides. Jf this is in fact a variable feature, then it would not on its own 
exclude Sardinioides from the ancestry of Aulopus. 
Thus it would appear that Sardinioides although not directly ancestral to Aulopus 
is extremely close to this ancestry. Both genera represent equivalent basal forms 
from which most, if not all, of the present day myctophiforms can be derived. 
The two genera Acrognathus and Nematonotus which were included in Woodward’s 
(1901) family Scopelidae both show strong affinities with the myctophiforms. The 
skull-roof of an undetermined species of Nematonotus (B.M.N.H. specimen P.48681) 
is shown in Text-figure 94, and it is very similar to those of both Sardinioides and 
Aulopus (Text-fig. 86). Apart from the skull-roof there are certain minor differences. 
The head and body of Nematonotus (Text-fig. 95, Nematonotus bottae) are somewhat 
deeper than in Sardinioides, and the premaxilla has a very long ascending process 
and a large articular process. One major difference occurs and this concerns the 
maxilla which in Nematonotus enters the gape behind the premaxilla, a feature 
not seen in any myctophiforms, and the maxilla is toothed. Corresponding to the 
entry of the maxilla into the gape, the bone itself is deeper and supports two supra- 
maxillae which are considerably larger than those in Sardinioides. Like Aulopus, 
but unlike Sardimioides, there is no gap between the second and third hypurals. 
The base of the first uroneural is forked in Nematonotus bottae (Text-fig. 96), this 
being a primitive feature seen in such forms as Elops (Text-fig. 80) and to a very 
slight extent in Salmo (Text-fig. 81), but in neither Sardinioides nor Aulopus (see 
also Patterson, 1968a : 226). 
