OP PISHES FEOM LAKE TANGANYIKA. 565 



some quite yoxmg specimens, 30 to 35 millim. long, from Mbete. Several specimens 

 had previously been obtained by Mr. Moore on the Usambara market ; but they were 

 badly preserved, and I took them to be the young of B. ferox. 



Pellegrin has placed Bathyhates in his division " Opisthospondyli " '. I can see no 

 justification for this course, the vertebral column not difi'ering notably from such forms 

 as Varatilapm rohusta and longiceps. Bathyhates is more nearly related to Paratilapia 

 than to any other genus, differing chiefly in the longer anal fin and in the dentition ; 

 this consists of an outer series of ankylosed teeth, behind v^^hich are one to three series 

 of hinged depressible teeth. We are now acquainted with three species, the principal 

 characters of which may be contrasted as follows : — 



I. Spinous dorsal fin not notcbed, the posterior spines subequal; 3 or 4 series of teeth in the 



upper jaw, 3 in the lower. 



D. XIV-XV 15-16; A. Ill 16-17; Sq. 78-100 ^^^^ 13-15 gill-rakers on 



lower part of anterior arch ; interocular width less than diameter of eye. B. ferox Blgr. 



D. XVI 16-17; A. Ill 17-18; Sq. 140-150 .,5^; 18-19 gill-rakers on 

 lower part of anterior arch ; interocular width greater than diameter of 

 eye B. fasciatus Blgr. 



II. Spinous dorsal notched, the last spine twice as long as the antepenultimate ; 2 or 8 series of 



teeth in the upper jaw, 2 in the lower. 

 D. XIII 15-16; A. Ill 16-17; Sq. 90-95^^; 12 gill-rakers on lower 

 part of anterior arch ; interocular width slightly less than diameter of 

 eye B. minor Blgr. 



I may here add, as a correction to the original diagnosis of the genus, that well- 

 preserved specimens show the maxillary bone not to be exposed when the mouth is 

 fully closed. 



Haplotaxodon, g. n. 



Mouth nearly vertical, directed upwards, with a single series of small, equal, conical 

 teeth. Maxillary exposed at the end. Body elongate, strongly compressed, covered 



^ So far as I am acquainted with the number of vertebrae in the African Cichlidee, the following species 

 only are to be referred to this division : — 



Enantiopus longianalis 144- 24 = 08. 



Xenotilapia sima 13 + 22 = 35. 



„ ornatipinnis 14-|- 24=38. 



Grammatotria lemairii 14 + 22 = 36. 



Trematocara unimaculatum 12 + 19^31. 



„ nigrifrons 13 + 16 = 29. 



Asprotilapia leptura 15 + 20 = 35. 



All generic forms restricted, according to the present state of our knowledge, to Lake Tanganyika. 



