416 Mr. G. A. Bouleno-er on 



L. — On the Species of Polypterus. 

 By G. A. Boulenger, F.R.S. 



In a paper published in the ' Denkschriften ' of the Vienna 

 Academy, 1881, Steindachner has shown that the three 

 species from the Nile, which had been well distinguished by 

 earlier authors and unfortunately united by Giinther, are 

 perfectly distinct, and the concise descriptions he has given 

 of Polypterus senegalus, Cuv., bichir, Geoffr., and End- 

 licherij Heck., accompanied by excellent figures, leave little 

 to be desired. The forms occurring in West Africa have 

 not been considered on that occasion, an omission the more 

 regrettable from the fact that they seem to bridge over 

 the differences so readily perceptible in dealing with the 

 Nile species. Having experienced some difficulty in the 

 determination of specimens from the Niger and the Congo, I 

 have found it necessary to review the whole material preserved 

 in the British Museum. This work has resulted in a more 

 rigid definition of the West-African species and in the addi- 

 tion of two new species from the Congo. Although, in spite 

 of Steindachner's observations, several authors have continued 

 to designate specimens under the collective name of P. bichir, 

 the specific distinction of the various forms cannot be ques- 

 tioned, for, even if there be a continuous chain connecting the 

 specimens with 5 dorsal spines with those with as many as 

 18, examination of large series shows the range of variation 

 in the number of spines never to exceed 4 in specimens other- 

 wise similar, the differences in these numbers being accom- 

 panied by others which cannot even be regarded as correlative. 

 Most of the species could be identified from specimens 

 stripped of their dorsal fin. For instance, the number of dorsal 

 spines is not correlated with the elongation of the body, nor 

 is this with the number of scales in the lateral line. The 

 head also furnishes important specific characters. 



In order to enable the reader to form a better idea of the 

 amount of variation within the limits of a species, I have 

 recorded the numbers of spines and scales and the proportions 

 of the head in all the specimens at my disposal. The figures 

 are arranged in four columns, which read as follows : — 



A. Number of spines in the dorsal tin. 



B. Number of scales between the upper extremity of the 



opercle and the middle rays of the caudal fin. 



C. Length of the head compared to its width in front of 



the opercular bones. 



D. Length of the head to the extremity of the opercular 



bone in the total length, caudal fin excluded. 



