ORGAN OP THE SILUROID HYPOPHTHALMUS. 113 



opercukim is of large size, the interoperculum A'ery small, and the suboperculum, as iisual, 

 absent. The preoperculum is fused with the hyomaudibular and quadrate, but a tartil- 

 aginous strip persists between the quadrate and the hyomaudibular. Part of the arch 

 dorsal to the quadrate is probably the metapterygoid, but there is uo suture to indicate 

 its separateness from the hyomaudibular. The palatine contrasts markedly in shape with 

 the rod-like palatine of Amiurus ; some cartilage persists at the articular knob which 

 rests in the fossa on the parethmoid. The edentulous dentary is of small size, the articu- 

 lar element, on the other hand, is large ; but uo others are to be detected in the adult. 



Although foreign to my present purpose, I have figured the singular shape of the 

 shoulder-girdle (Fig. 5.) the clavicles being prolonged as far as the symphysis of the lower 

 jaw . 



The series of selected sections, represented in. Figs. 6-16, was made from one of the 

 specimens at my disposal, with the view of elucidating those relationships of the auditory 

 organ to the air-bladder, which, it would have been impossible to arrive at satisfactorily, 

 even with the most careful dissection of a single specimen. The series extended from 

 behind the trigeminal foramen to the plane of emergence of the sixth spinal nerve, and 

 included over 300 sections. 



Fig. 6 represents one of the foremost of these ; it passes through the posterior cranial 

 fontanelle, the supraoccipital (parietal element), spheuotic, prootir, l)asi,sphenoid (and para- 

 sphenoid). Some cartilage persists between the supraoccipital and sphenotic and in the 

 sphenotico-prootic suture. The brain is evidently similar to that of Amiurus (cf. my Fig. 10, 

 PI. y, for cit.). From the facio-trigeminal ganglion there is given off a branch to the neuro- 

 mastic canal in the spheuotic, which I have identified as the ramus oticus in Amiurus 

 (No. 9, p. 366) and Lepidosteus (No. 10, p. 490). In the skin overlying the sphenotic are 

 branches of the neuromastic canal, which, however, do not lontain neuromasts, but are 

 the complex channels by which the canal puts itself in communication with the outside. 

 Solger has referred to this method of the opening of the neuromastic canals in the genus 

 at present under consideration (No. T, p. 3'72). 



Fig. 7 hardly requires a separate description ; it falls through the plane of escape of 

 the hyomaudibular and palatine branches of the facial, and indicates how the prootic and 

 sphenotic participate in the formation of the fossa which conceals the anterior bend of the 

 anterior semicircular canal. Although all the bones of the skull are fragile, this is espe- 

 cially true of those of the roof which have an extremely spongy texture, the intervals 

 between the delicate osseous trabeculœ being occupied largely by fat cells. 



The section represented in Fig. 8 falls through the plane of the sphenotico-pterotic 

 suture, where a considerable amount of cartilage persists, aud even enters into the forma- 

 tion of the hyomandibular fossa. It will be seeu that the neuromastic canal lies directly 

 on the surface of the cartilage, or is separated therefi-om only by a very delicate lamina of 

 bone. As one would expect from the great development of the system of neuromastic 

 canals in this fish, the tuberculum acusticum, that region of the brain iu which the dorsal 

 branches of the various cranial nerves centre, is of unusual dimensions. 



Our next section (Fig. 9) is entirely behind the prootic, and thus at a considerable 

 interval from the last. It falls through the plane of the utriculus proper, and near the 

 ductus sacculo-utricularis, which eiFects the communication between the upper and lower 



Sec. IV., 1885. 15. 



