NOTES ON MEDIAN AND PAIRED FINS OP FISH. 351 



of segments may share in the formation of the girdle, and 

 that when no diazonal nerves are present only one cartilaginous 

 segmental element is fully developed, at all events at the 

 point where the nerves pass outwards to the limb. When 

 several nerves pass through the same foramen we may 

 suppose that the cartilaginous elements between them have 

 been suppressed. It is interesting to note that in the case of 

 the Chondrostei (Thacher 35, Wiedersheim 36, Mollier 26) the 

 pelvic girdle still shows distinct traces of segmentation. Since, 

 however, the girdles are structures which grow inwards, en- 

 veloping the nerve-plexus, with which they only come into 

 secondary connection, it is quite possible that all strict 

 metameric concordance between the two has been modified or 

 lost in most cases. But a limb-girdle may be trans- 

 posed, like a plexus, by the addition of new elements at one 

 end and their disappearance at the opposite end. And thus 

 is brought about the apparent backAvard or forward motion 

 of a girdle through a number of segmental nerves, or, in other 

 words, the passage of nerves through a girdle. 



To this theory of transposition it may be objected that, if 

 true, the limbs and girdles of the Gnathostomata are not 

 strictly homologous. Now, if by the homology of two structures 

 we mean tliat they are produced by the same number of 

 segments, occupying in both cases the same place in the 

 metameric series, the limbs and girdles are certainly not 

 always homologous. In this strict and narrow sense they are 

 often not homologous amongst closely allied species, nor in 

 individuals of the same species, nor even on the two sides of 

 the same individual. Fiirbringer, Brans, and Punnett have 

 clearly demonstrated the great variability of the nerve-plexus 

 supplying the paired limbs. So long as a distinct individuality 

 and persistence are attributed to each segment, so long as 

 segment x of one animal is considered to be represented 

 only by the same segment x in another animal, the term 

 ''homology" can only be applied in a general sense 

 to the limb and its nerve-plexus, etc., as a whole. And let it 

 not be imagined that we can escape from this conclusion by 



