PEOP. E. EAT LANKESTEE ON LEPIDOSIEEN AND PEOTOPTEETJS. 21 



so as to expose the anus on the left side of the animal's body, and showing 

 the villi (found only in the male) on the dorso-mediad face of the limb. 

 Drawn by J. Bayzand : twice the natural size, linear. 



Fig. 4. Scale from middle of the side of the body of Ceratodus forsteri, Krefft. 

 X 3 diam. h, dotted line marking off the exposed portion of the scale (to 

 the right) from that which is sunk beneath the surface. 



Fig. 5. Scale from a similar position from Protopterus annectens. X 3 diam. 



Fig. 6. Scale from a similar position from the Paraguayan Lepidosiren, X 3 diam. 

 For explanation of reference-letters a and h see footnote. 



Fig. 7. Denticulation of the scale of Ceratodus. X 180 diam. 



Fig. 8. Denticulation of the scale oi Protopterus. X 180 diam. 



Fig. 9. Denticulation of the scale oi Lepidosiren. X 180 diam. 



Postscript. — July 8#A, 1895. 



Since the foregoing paper was written and the figures of the Plate drawn, I have made 

 it my business to examine as far as possible the specimens of Lepidosiren from the 

 Amazons which exist in European Museums. 



I visited Paris in April, and by the courtesy of the Director of the Jardin des Plantes, 

 M. Milne-Edwards, and of Prof. Le Vaillant was permitted to] examine carefully and to 

 measure Castelnau's specimen, as also a fragmentary skin from an old Portuguese 

 collection, which is probably referable to the Amazonian Lepidosiren. 



I proceeded to Florence, where my old friend Prof. Giglioli placed at my service — 

 for the purpose of determining the justification of Ehlers's proposal to form the new 

 species, Lepidosiren articulata — his two specimens from the Amazons. 



Lastly, I wrote to Dr. Steindachner, of Vienna, to ask him to examine Natterer's two 

 original specimens in order to determine whether the skeletal axis of the limb is 

 " articulated," or a simple continuous rod of cartilage. Dr. Steindachner wrote to 

 inform me that the hinder limbs of the specimen anatomized by BischoflF appear to 

 be lost : they cannot be found. At the same time he had the great kindness to send to 

 me Natterer's untouched second specimen in order that I might examine it myself 

 I have examined it, and returned it to Dr. Steindachner a month ago. 



Thus I have myself examined every specimen of Brazilian Lepidosiren known to 



and of the Dipnoi as a natural group, appears to be that the scales of Protopterus and Lepidosiren are in a state 

 of retrogressive modification or degeneration as compared with those of Ceratodus. It would seem that, 

 whereas the ancestral form presents a freely exposed posterior margin to the scale, the whole scale has shrunk 

 and withdrawn, so to speak, within and beneath the outermost layers of the dermis in Protoptena and Lqndo- 

 siren, whilst Ceratodus retains the ancestral character. Of course, behind and beyond that (in historical 

 sequence) there is a hypothetical series of more primitive ancestral phases which lead to the multitudinous 

 dermic denticles of primitive fishes. 



