187(3.] PROF. T. H. HUXLEY ON CERATODUS FORSTERI. 47 



the last. A more important point is that the second shows no trace 

 of such divisions as those described by Dr. Gunther. To make 

 sure of this I made a thin microscopic section of this cartilage on 

 the right side, and thereby satisfied myself of the homogeneity of the 

 cartilage of which it is composed. 



I find no true joint between the proximal median piece and the 

 scapular arch, the connexion between the two being effected by a 

 solid fibrous mass. 



Again, the "slight irregularities" in the distribution of the rays, 

 in respect of the median pieces, of which Dr. Gunther speaks, are 

 constant peculiarities of no small importance. This becomes obvious 

 when the fin of Ceratodus is compared with that of other fishes. 

 Hut before proceeding to this point I must make a few remarks on 

 the normal and primitive position of the vertebrate limb, and on the 

 changes from that normal position which take place in fishes on the 

 one hand, and the higher Vertebrata on the other, as, for want of 

 attending to this fundamental matter, grave errors have crept into 

 the interpretation of the parts of the limbs of different vertebrates. 



The fins occupy the normal position in such a fish as Scyllium. 

 When the axis of the body is horizontal the plane of the fin is also 

 horizontal. Its upper surface is covered by a continuation of the 

 dorsal integument, and its lower surface by that of the ventral side 

 of the body. The distinctive spotting of the dorsal aspect in the 

 Dogfish makes this very plain ; therefore, to adopt the nomenclature 

 which I some time ago proposed, the convex thick edge of the fin is 

 prseaxial, while its concavo-convex thin edge is postaxial, and its 

 two aspects look respectively upwards and downwards. 



In Ceratodus, however, the fin has undergone a rotation upon its 

 axis, in virtue of which its proper ventral surface looks more or less 

 outwards, and its proper dorsal surface more or less inwards ; and at 

 the same time the prseaxial edge is turned upwards, while the post- 

 axial edge is turned downwards. This is very marked when the fin 

 is applied to the trunk ; but the primitive disposition of the surfaces 

 and edges of the limb becomes obvious when the fin is made to stand 

 out at right angles to the axis of the body. 



In Acipenser, as in any Teleostean fish, the rotation becomes still 

 more marked ; and divarication of the fin does not greatly diminish 

 it, by reason of the articulation of the praeaxial fin-ray with the pec- 

 toral arch at a point higher than the proper glenoid cavity. 



When Ceratodus, or a Teleostean fish, is placed side by side with 

 a Urodele Amphibian, such as Menobranchus, in which the fore limb 

 naturally takes a position nearly parallel with the axis of the trunk, 

 the manus being directed backwards, nothing is more natural than 

 to imagine that the outer and inner aspects and the dorsal and 

 ventral edges of the two correspond. 



But a very little consideration will show that nothing can be more 

 erroneous ; for the outer surface of the Menobranchus limb is its 

 proper dorsal aspect, and the inner surface is its ventral aspect, as 

 will be seen the moment the limb is placed at right angles with the 

 trunk. In fact, though in the amphibian, as in the fish, the limb 



