542 n '^- i^'^r.T.ARi), 



reniarkablo fact is schmi , that they a^Tce closely. The bearing of 

 this will be evident when it is remembered how far removed the South 

 American and African faunas are and how ancient the former is. 

 Chneiosionius , also belongs to that subfamily of Siluroids which 

 l)0ssibly contains the most archaic representatives of all the Teleostei, 

 forms in which a protective armour is present consisting of scutes 

 which are little more than aggregations of true dermal teeth and 

 dermal teeth also which agree in essential characters with the teeth 

 of the oral cavity. (The importance of this last point has been in- 

 dicated by Heutwig.) We learn thus that the mucous canals of 

 Clarias and Auchenaspis represent a very early type. 



Making the comparison now in detail and commencing with the 

 infraorbital system it will be seen that the latter is much more 

 elongated in Chaetostomus than in Clarias, Auchenaspis to a certain 

 extent however lying between the two. It seems probable that the 

 first pore in Chaetostomus is not represented in the others. The rest 

 of the pores however correspond. 



Pore 6 in Chaetostomus and 5 in Auchenaspis does not as pore h 

 in Clarias stand at the end of a branch of considerable morphological 

 importance but is like an ordinary pore. Pores 7 in Chaetostomus 

 and 6 in the other two are situated at the ends of considerable 

 branches. In all three forms there is no sense organ between this pore 

 and the junction of supra- and infraorbital systems. Correspondingly 

 there is no pore at that junction. 



In the supraorbital system the sense organs correspond exactly 

 save for the already mentioned extra one no. 5 in Clarias. The 

 branch to pore 3 in all bends backwards and outwards behind the 

 nose. The most striking feature is the distribution of the branch which 

 proceed medianly between sense organs 3 and 4. In Chaetostomus 

 it forms a well developed frontal commissure, in 

 Clarias it is a complete commissure but one becoming 

 somewhat rudimentary, in Auchenaspis it is no longer 

 a commissure, but opens with a dichotomous division at 2 pores. 



The median branch containing sense organ 5 (6 in Clarias) corre- 

 sponds in all three being however most developed in Clarias. 



In the main canal it may be noticed that in all three forms, the 

 opercular canal is in complete communication with the body line and 

 the branch extending towards the shouldergirdle is 

 present. The interesting branch extending to -^ov^ ^ m Auchenaspis 

 is indicated only in Chaetostomus. 



