24 PEOF. E. EAT LANKESTEE ON LEPIDOSIEEN AND PEOTOPTEEUS. 



has or has not. The words which he uses do not amount to a statement to the effect 

 that he has. He says that five specimens out of thirty-two submitted to him by 

 Dr. Bohls from Paraguay differed in certain features of the head from the other 

 twenty-seven ; further, that one (or more ]) of the twenty-seven had segmented limb- 

 axes, and therefore (it seems) he concludes (but does not state as the result of 

 observation) that the remaining five must be referred to Fitzinger's L. paradoxa, and 

 must have an unsegmented limb-axis, since Bischoff figured such an axis in one of the 

 original specimens of L. paradoxa. The probability, from my observations, is that 

 Bischoff erroneously omitted the segmentation of the limb-axis in his figure, and did 

 not look into the matter carefully. It seems highly probable that all Brazilian 

 specimens of Lepidosiren, and all Paraguayan specimens of Lepidosiren, have, like 

 the African Protopterus, a segmented limb-axis. The chief ground put forward by 

 Ehlers for erecting a new species, " Lepidosiren articulata" from Paraguay, would, if 

 my surmise be correct, fall to the ground. 



Whether any specific distinction between Paraguayan and Amazonian specimens of 

 Lepidosiren can be based upon other grounds remains to be seen. The characters 

 derived from small differences in the shape of the head, relied upon by Ehlers, cannot, 

 in my opinion, be considered as of any value in specimens preserved in spirit, and 

 pressed down in jars or travelling-cases. The most striking and uniform difference 

 which I have observed is in colour : the Paraguayan specimens are black, the Brazilian 

 specimens are clear brown. There is, however, little doubt that this is largely, if 

 not altogether, due to the longer action of alcohol and sunlight upon the Brazilian 

 specimens. 



Lastly, we may compare the measurements of the Brazilian and Paraguayan specimens. 



If we write down, for the three Brazilian specimens, the measurements already given 

 in terms of head-length, it appears that Castelnau's specimen has a total length of 

 9*54 units, an inter-membral length of 5 - 5, a post-membral of 3, and a cervico-dorsal 

 of 3 ; for Giglioli's large specimen the corresponding figures are — 10, 5 - 7, 3"4, and 

 3'63 ; for the Vienna specimen they are — 10'06, 5-64, 4, and 3. If we now take the 

 average of these three Brazilian specimens, and compare it with the average of the five 

 Paraguayan specimens given on p. 15, we obtain the following striking result: — 



Total length. Inter-rnembral. Post-membral. Cervico-dorsal. 



Brazilian average in head-length units 9-87 5-25 3 - 5 3 - 21 



Paraguayan average in head-length units .... 9'75 5-21 3-59 3"2 



The specimens in both series were taken altogether by hazard, yet there is the closest 

 agreement. No d|.oubt such a result does not exclude the possibility of a specific 

 difference being discovered separating the Paraguayan from the Brazilian Lepidosiren^ 

 But it renders the! existence of such a difference very unlikely, and I have already 

 shown that it is extremely probable that the character relied on by Prof. Ehlers in 

 order to separate soine Paraguayan specimens from the Brazilian species has no existence 

 in fact. 



