﻿31 69 



expedient tliat the imisculaliire should for the most part move out to the inner 

 surface of the abdominal armour''. 



The ventral fins are coalesced along their inner margin as in the Gobies. 

 The number of rays is 5, 1 spinous and 4 soft, but unarticulated and unbran- 

 ched rays; these naturally are distinctly composed of paired parts in contrast to 

 the spinous ray. The spinous ray is very fine, much shorter than the other rays, 

 sometimes of ditTerent length on the two sides (in a 5 of A.strigata I (ind it half 

 as long on the left as on the right side); it may be so small, half rudimentary 

 that il is difficult to find; this is the condition I find in 2 A. punctulata. In A. 

 strigata the males have long ventral fins the two posterior rays especially being 

 long; the ventral fins, which are placed on the 8th ventral plate, reach in this 

 case to or beyond the boundary between the 9th and 10th plate, sometimes to the 

 middle of the latter or even to its posterior border. Whether there is a sexual 

 dilTerence in the ventral fins in the other two species I am unable to determine; 

 I do not think so however and in any case it will scarcely be obvious. In 10 

 specimens of A. scutata before me the variation in length of the ventrals is quite 

 unimportant and in 11 A. punctulata, which I have examined with regard to this 

 point, I likewise find but little variation; I find that the fin never reaches beyond 

 the 9th ventral plate. That specimens of A.strigata with long ventrals are really 

 S, as stated by earlier authors, I have had the opportunity to confirm by the 

 examination of one specimen'. 



Remaining anatomical features. 



With regard to the anatomy of the soft parts I shall restrict myself to some 

 few remarks, as my investigation was chiefly concerned with the osteology; the 

 remaining organs have only received occasional attention. 



The character of the musculature is in high degree determined by the stiff 

 armour. The lateral muscles have for a great part disappeared; only the dorsal 

 portion is fully preserved. The part of this which lies on each side along the 

 elongated, anterior vertebræ, above the attachment of these to the ridges on the 

 lateral plates, are attached in front to the skull, very firmly especially to the 

 supraoccipital, by means of a thick tendinous mass, which further back divides 

 into two long tendinous strips along the upper border of the muscles; the whole 

 of this muscular part probably represents a number of muscular segments corre- 

 sponding to the 5 — 6 vertebræ and we find also, that it is penetrated longitudinally 

 by ca. 10 long, shining tendinous strips, which seem parallel but in reality con- 

 verge posteriorly at very pointed angles, whose apices lie near the limit of the 

 movable part of the spine and somewhat further back. The muscular mass is 

 specially tendinous along the ventral edge of this region and constantly more ten- 

 dinous the more we approach the movable part of the vertebral column ; the 

 tendons are fixed to the transverse processes of the posterior, movable abdominal 

 vertebræ and those of the anterior caudal vertebræ. The dorso-lateral muscles, 



