340 THE TANGANYIKA PROBLEM. 



ganyika, some of the Charicinidce, and most of the 

 CichlidcB, as in all probability representing the now 

 scattered piscine portion of the halolimnic fauna. 



In approaching this matter we must, as I have insisted 

 already, bear in mind that most fishes are active, migratory 

 animals, and that, once established in the fresh-waters of 

 a continent at any one point, are sure to rapidly spread 

 through the water-systems of the interior, which in all 

 great land-masses are often, temporarily, more or less 

 connected together. Yet, in spite of this, as I have 

 shown in Chapter VIII., more than half the Old World's 

 species of Cichlidae are peculiar to Tanganyika at the 

 present time. 



Palaeontology is as yet silent with respect to the origin 

 of this group ; but it is more than probable that its 

 forerunners were once widely spread in the sea, and 

 from these facts it would seem that we must draw a 

 conclusion similar to the inference which we might draw i 

 from the distribution of blow-fiies we knew r to have 

 emanated from maggots in a piece of rotten meat in one 

 of the rooms of a house. If we found a few blow-flies in 

 one of the rooms of our supposed house, and in another 

 more, and in another a whole swarm, we should infer that 

 the rotten meat was in that particular room. And in the 

 case of the Cichlid fishes, believing their forerunners to 

 have been once in the sea, but finding that in the Old 

 World they are now restricted to Tanganyika, and are 

 distributed more and more sparsely as we radiate over 

 the Old World from this lake, we must infer that the 

 sea in which they used to exist must have been in the 

 neighbourhood of Tanganyika. 



Considering the distribution of the families of fish which 

 are now present in Tanganyika, they are all more or less 



