102 REPOKT— 1889. 



The head of Dacbjlopferus voUtans is superficially covered by a com- 

 plete bony layer. In the natural condition this shows no appearance of 

 satares, and is prodaced backwards, dorsally, beyond the region of the 

 skall. Here it bifarcates, each branch forming a lai-ge flat plate resting 

 on what is popularly known as the ' shoulder,' and each ending in a 

 flattened spine. For convenience I have called these the dorsal plates. 

 After maceration in a very strong solution of caustic potash this layer is 

 separable into distinct areas, and can be removed, leaving the true skull 

 bones exposed beneath. It is then seen to be a true sheath arranged 

 with a view to strengrth. 



Each superficial suture is above the body of a true bone, and each 

 deep suture in the skull proper is below the body of one of these areas. 



Of the skull itself tbe occipital arch is the only one deserving of 

 special attention. It is peculiar, but will be better understood after 

 examining the swimming-bladder itself. The vertebral column is 

 already well known to be peculiar. As in Fistidaria, a Japanese pipe- 

 fish, the first four vertebrce have coalesced so as to form a rigid tube ; 

 their neural spines have also united to form a vertical plate. I can gain 

 no information concerning Fistularia, but in IDactylopterus it appears to 

 me that this peculiarity is subservient to the uses of the swimming- 

 bladder, just as I think the dorsal plates are. Both give rigidity and 

 strength to this segment of the fish. 



The swimming-bladder, so far as I am aware, has a unique position, 

 inasmuch as it is not situated below, but above, the vertebral column, not 

 forming part of the abdominal contents, but situated dorsally in a special 

 cavity of its own. When the abdominal cavity is opened ventrally, and the 

 viscera removed, only the ventral surface of the bladder is seen, forming 

 part of the dorsal boundary of the cavity. Seen from this point of view, 

 it is formed of a broad central portion, white and tendinous, and of two 

 lateral portions, strongly muscular. 



The kidneys overlap the swimming bladder posteriorly, and the oeso- 

 phagus and pericardial cavity partially obscure its anterior end. 



On removing the abdominal walls so as to obtain a view of the side, 

 and also of the back, the lateral muscles are seen to continue upwards, to 

 curve inwards towards the median line, to be reflected downwai'ds on 

 each side of the vertical neural plate, and finally to become attached to 

 the bodies of the vertebrEe, whose spines go to form that plate. 



Leydig ' gives a short description of the swimming-bladder of 

 Dacti/lopteruD, the result probably of a somewhat cursory observation, as 

 he seems to have had the impression that there was only one large muscle 

 covering the dorsum of the bladder. But, as the four coalesced vertebrae 

 of the spinal column necessarily intervene, it will be at once evident that 

 the bladder is here partially divided into equal portions, forming, as will 

 be seen later, its two primary divisions. Giinther ^ recognises the bladder 

 as being divided into ' two halves, each having a large muscle.' A 

 lateral view of one of those muscles gives an elliptical ontline, the long 

 axis running, from the posterior end, forwards and slightly upwards. 

 Tracing, however, the dorsal surface forwards, and on its inner or median 

 aspect, we find that here there is, on each side, a secondary portion con- 

 cealed immediately below the dorsal plate before mentioned. Carefully 

 scraping away this plate, it is seen to be composed of an extremely thin 



Lehrhuch der Histologie. = Introduction to tlie Study of FiiJies. 



i 



