﻿51 89 



arch and the cristæ inferiores e of the two sides, so that they pass out to the 

 fin-iays through the opening x in fig. 31. The long horizontal process /) corre- 

 sponds to the one which projects from the corresponding position in the above- 

 mentioned spinous-rayed lishes, short in Sebastes and the Colloids, long in Trigla, 

 and here also lies in the musculature of the abdominal wall; hut in Centriscus 

 it has obtained a much more considerable size in comparison with the true pubic 

 bones, corresponding to the new function of covering the ventral (ins; in addition 

 to this, the crista superior has become connected with it and contributes to its 

 formation, thus producing the hollow h. The articulating surface for the rays of 

 the ventral fin lies in front of this hollow, at y in fig. 31. 



Contpared with Amphisile, it is evident, that the pubic arch in the latter 

 corresponds to the main portion of the arch in Centriscus, the process p not being 

 developed at all. In both genera the arch is directed upwards in a similar manner, 

 but the much greater compression of the abdomen in Amphisile has influenced 

 the pubic arch to a much greater extent both with regard to form and position i*. 



The number of rays in the ventral fins is 5; the outermost ray is a 

 perfectly typical spinous ray, shorter than the others and without the covering of 

 rows of dermal teeth (scales), which are characteristic of the remaining rays, 

 especially on the proximal part; the spinous ray may be pressed close up against 

 the considerably longer 2nd ray and is thus easily overlooked. The other four 

 rays are double, articulated and dichotomously branched. The ventral fins are 

 situated quite close together but do not seem to be fused'^. 



Remaining anatomical features. 

 Musculature. The stiffness of the anterior portion of the trunk also has 

 a certain influence in Centriscus. Of the dorsal part of the lateral muscles of the 

 trunk the portion lying under the dorsal armour, along the elongated, immovable 

 vertebræ, shows some amount of transformation, recalling that in Amphisile. The 

 whole of this has in fact become one mass, in which the segmentation can only 

 be seen with difficulty; towards the back part only can the myomeres be seen. 

 In front this muscular mass is firmly attached to the postoccipital surface of the 

 skull, especially above by means of strong tendinous ends to the ridge along the 

 side of the supraoccipital; the lower margin further back is strongly tendinous and 

 the tendons run back to become attached to the transverse processes of the movable 

 abdominal vertebræ. The remaining part of the dorso-lateral musculature is seg- 

 mented as usual, likewise the ventral portion on the tail and the sides of the 

 body; but a part of the latter is extremely thin and becomes weaker and thinner 

 forwards; all musculature is lacking on a triangular area between the broad up[)er 

 portion of the clavicle, at the base of the postclavicle, and the weak intermuscular 

 ligament which springs from the transverse process of the 4th vertebra; the area 

 in question is partly covered over by the portion of the dorsal armour which lies 

 below the ridge and corresponds to the non-muscular area in Amphisile through 



12- 



