TRANSACTIONS OF SECTION D. 737 



In Tetrodon kispidus, immaculatus and nigropunctatus, a tentacle-like process of 

 the skin is present on either side of the head, between the snout and the eye, the 

 base of which is surrounded by a narrow circular fold, while distally it becomes 

 forked so as to form two broad divergent lamellae. The surfaces of these two 

 lamellae which are turned towards one another, are pigmented, and are covered 

 over by a network of delicate ridges, in the meshes of which the sensory organs 

 lie. The outer surfaces of the lamellae are smooth. 



Somewhat behind the middle line of the internal wall of the orbit, the exceed- 

 ingly dehcate olfactory nerve passes out of the skull-wall. Directly after its exit 

 it becomes surrounded by a thick fibrous sheath, which protects it from the move- 

 ments of the eyeball and its muscles. 



After passing over the superior oblique muscle, the nerve comes to lie dorsally 

 to the well-developed muscles of the jaws, being closely jammed in between these 

 and the wall of the skidl. At the same time it is protected from the pressure of the 

 masticatory muscles by a dense fibrous plate. 



From this point, instead of passing to the base of an olfactory sac, it extends 

 directly outwards into the skin, and thence into the above-described cylindrical 

 nasal process ; within this it branches out iut» a series of twigs arrano-ed in a 

 circle around the longitudinal axis of the process. 



These nerve-twigs extend throughout each lamella, and pass outwards into the 

 sensory cells, the arrangement of which is exactly similar to that of segmental 

 sense organs in the skin of fishes. 



The olfactory organ of Tetrodon papua is of special interest, inasmuch as it has 

 undergone a very considerable degeneration. No trace of projecting nasal folds 

 can be seen, and for some time I thought that a nasal organ was quite wantino- in 

 this fish. Indeed, from a physiological point of view, this is very likely the case, 

 but if the brain be examined, the olfactory nerves can be distinguished by the aid 

 of a lens. They have the form of hair-like threads, the relations of which to the 

 orbit and to the muscles of the jaws are quite similar to those I have described 

 above in the case of other Tetrodonts. Instead, however, of branching out into a 

 process of the skin, they end on a level with the general surface of the integument, 

 which shows at this point a small pigment-spot. If I had not carefully followed 

 out the course of the olfactory nerve, this spot would have escaped my notice, for 

 the skin of the head is provided with numerous similar pigment-spots in this 

 region. I cannot state with certainty whether a neuro-epithelium was present, as 

 the specimen was not sufficiently well preserved for histological examination. 



Besides the species of Tetrodon already named, I have also examined Tetrodon 

 pardalis and Diodon maculatus. In both of these, nasal processes are also present 

 having the form of blunt cones. Instead, however, of being solid, they are hollow, 

 the enclosed cavity communicating with the exterior by two apertures, throuf'h 

 which a current of water can pass as the fish moves about. 



• The lining-wall of the cavity is raised into a number of fold-like processes, the 

 arrangement of which is similar to that of the valves in the conus arteriosus of 

 certain fishes ; in this way a large surface is produced for the sensory organs. 



As in the forms already described, a proper olfactory sac is wantino-, and the 

 olfactory nerve passes directly towards the outer skin, finally ending in the nasal 

 processes. 



These observations lead me to the following conclusions : — 

 The peculiar structure of the olfactory organ in the genus Tetrodon cannot be 

 looked upon as primitive, but must have arisen secondarily. Tetrodonts must 

 formerly have possessed a proper olfactory sac, more or less deeply sunk into the 

 skull. Moreover, this sac was provided with a membranous tube'leading to the 

 exterior, similar to that present in Muraenoids, Polypterus, and many other fishes. 

 In the course of phylogenetic development, as the Tetrodonts began to browse 

 upon corals and hard shells of mollusks, the masticatory muscles must have under- 

 gone a correspondingly strong development. In consequence of this, the points of 

 origin of these muscles extended further and further over the anterior part of the 

 skull, and passed upwards between the snout and eyes. Thus they gradually 

 replaced the olfactory sac, which was formerly present in this region : while the 

 1»S'^ 3 B 



