HEADS AND TAILS. 53 



fishes, such as the flounder, the salmon (fig. 6, (7.), 

 and the perch, for instance. In this we meet again, 

 apparently, with the same perfect symmetry that 

 characterised the diphycercal tail of Frotopterus 

 and some primitive sharks. This apparent sym- 

 metry has been arrived at by some exceedingly 

 interesting stages, fraught with a deep significance 

 when we come to look below the surface. The 

 lessons which these stages have to teach we will 

 now proceed to discuss. 



In our investigation we must begin with the 

 larval fish — that is to say, with a very young 

 fish, just before, or soon after, it has left the egg. 

 The tail of such a fish — that of a young flounder, 

 for instance — is, we shall find, truly diphycercal. 

 This we will call stage No. 1. In stage No. 2 

 the axis of the tail — i.e. the end of the vertebral 

 column, beghis, though ever so slightly, to turn 

 upwards, and from its lower surface numerous 

 rod-shaped processes are beginning to make their 

 appearance (fig. 6, D.). Our stage No. 3 ex- 

 hibits the tail in a bi-lobed form (fig. 6, E.). 

 The upper lobe is developed around the extreme 

 end of the axis of the tail, the lower from its 

 ventral surface. Passing to stage 4, we notice 

 the upper lobe has undergone a great decline, 

 whilst the lower has relatively increased in 

 size. In stage 5 the disproportion between the 

 two is enormous, the upper lobe having almost 

 entirely disappeared. In stage 6, our last, 

 the dorsal lobe is barely traceable, whilst the 

 ventral lobe has come to assume a superficially 

 perfect symmetry (fig. 6, F,). Thus, in the 

 life-history of a single fish, all three forms of tail 



