PAEATILAPIA. 473 



Paratilapia serranus was described by Prof. Pfeffer from a single specimen obtained 

 by Dr. Stuhlmann at Bukoba, a German station on Lake Victoria at about 1° 21' S. lat. 

 Examples have since been obtained by M. Alluaud in Kavirondo Bay, and I have 

 before me the following series : — 



4 Lake Victoria. — Sir Harry Johnston, 1901. 



2 Lake Victoria.— Col. Delme Kadcliffe, 1903. 



20 Entebbe.— Degen, 14-23.6.05. 



10 Bnnjako.— Degen, 5-14.11.05. 



10 Nsonga (Bussi).— Degen, 3.11.05. 



1 Buganga. — Degen, 22.11.05. 



A female specimen 140 millim. long, from Bunjako, in Mr. Degen's collection, has 

 the mouth and pharynx full of young in an advanced condition, 16 millim. long, but 

 with the yolk-sac not quite absorbed. I have not been able to count the number 

 of young, not wishing to remove them all, but there seems to be about thirty. 



Paratilapia vittata, Blgr., from Lake Kivu, is very closely related to P. serranus 

 and P. prognatha. These three species should perhaps be united. 



4. PAEATILAPIA GUIABTI. 

 (Plate LXXXVIL fig. 2.) 

 Tilapia guiarti, Pellegrin, Mem. Soc. Zool. France, xvii. 1905, p. 184, pi. xvi. fig. 1. 



This fish, originally described from a single young specimen from Kavirondo Bay 

 (Coll. Alluaud), is so closely allied to P. serranus that it is with some hesitation that 

 I regard it as a distinct species. The teeth are very similar to those of P. serranus, 

 those of the outer series being bicuspid in the young and unicuspid in the adult, but 

 the rows behind them are composed exclusively of strongly tricuspid teeth. The mouth 

 is smaller than in P. serranus, not quite reaching to below the anterior border of the 

 eye, and the caudal fin. truncate in the young, is feebly but distinctly emarginate in 

 the adult. 



Coloration and sexual differences as in P. serranus, but dark longitudinal bands 

 constantly absent. 



I have examined about twenty-five specimens from Entebbe, Lake Victoria, 

 14-23.6.05, in Mr. Degen's Collection. 



The following table of measurements, &c, shows the proportions to vary quite as 

 much as in P. serranus. 



3p 



