ESPECIALLY MYCTOPHOIDS 177 
genera are also predatory as evidenced by the well-developed dentition on all the 
jaw bones. The jaw suspensorium remains vertical in position irrespective of the 
lengthening of the jaws and is always shallow. This factor alone tends to indicate 
that these genera moved on to prey slowly and undetected. A fast movement 
would be hazardous to the long thin jaws which would have had to have been opened 
during this rush. 
Bodily elongation occurs within the Dercetidae as a result basically of the multi- 
plication of vertebral elements (as many as 80 being present in Dercetis gracilis, 
Text-fig. 25). Additional lengthening is achieved by extension of the vertebral 
centra in the anterior precaudal region in both Dercetis and Rhynchodercetis. The 
outcome of this extension of the centra is a fish with a very long and narrow ‘ neck’ 
region, well shown in Dercetis rostralis (Text-fig. 26), in which the anterior centra are 
as much as six times as long as they are deep. Together with the extension of the 
“neck ’ there is the production of enlarged transverse processes projecting strongly 
laterally (Text-fig. 30A). Siegfried (1966 : 214) has compared the dercetids with the 
Gasterosteiformes in respect of the vertebral composition. Elongation of the 
anterior precaudal vertebrae occurs in the Aulostomoidei (Awlostoma) and the 
Syngnathoidei (Syngnathus), and in these genera large transverse processes occur. 
The gasterosteiforms are in no way related to the dercetids, since the elongate 
dercetid jaw is at complete variance with the very small jaws at the end of a long 
tubular snout encountered in the Gasterosteiformes. 
In Rhynchodercetis hakelensis (Text-fig. 28) long, linearly arranged ossifications in 
the mid-dorsal line between the occiput and the origin of the dorsal fin may represent 
ossifications within a median dorsal ligament. If the transverse processes were in 
life connected one to another by ligaments, then together with the ossified dorsal 
ligament, the ‘neck’ could have been held rigidly in a straight line. Since the 
median fins occur in the posterior half of the body and the caudal region is relatively 
stout, then this may support the assumption that these fish moved slowly and 
stealthily towards their prey, the anterior body being held rigid and thus creating 
little turbulence in the water, whilst the main propulsive force is concentrated at 
the extreme posterior end of the body. On the other hand the modifications of the 
body may have enabled the fish to make a quick darting movement, this however 
has been shown not to be in accord with the jaw structure. Arambourg (1954 : I1Q) 
stated that Rhynchodercetis yovanovitcht resembled the Scomberesocidae and certain 
Xiphiidae, and was probably exclusively pelagic. This would seem to be a reason- 
able conclusion and the habit of the gar-fish, Belone, is closest to that of the dercetids. 
Squamation in the Dercetidae is reduced to several isolated scute rows along the 
flanks, the major rows being composed of scales conforming to a basic triradiate plan 
(Text-figs. 82A—D). In general, reduction in squamation is correlated with the 
presence of a distensible stomach, and this is so in the dercetids, since several 
specimens examined contained whole fish within the body cavity (see also Woodward, 
1942a : 555). 
The genus Prvionolepis has so far not been mentioned in the systematic descriptions 
or in the discussion. This genus, originally refered to the Enchodontidae by 
