The Lateral Line Systems in Siluroids. 547 



sider the fused frontal and postfrontal bones in which it joins the 

 supraorbital system. 



The supraorbital groove commences at the anterior edge of the 

 prefrontal which it traverses passing then into the frontal where it 

 meets the infraorbital system. In the frontal there is a fully 

 developed frontal transverse commissure as in Chaeto- 

 stonius or in Clarias. (The existence of this commissure is also re- 

 corded by Bodenstein in Coitus and Guitel in Lepadogaster and 

 Gohiesox. It may be not infrequent in Teleostei.) 



With regard to the continuation of the groove posteriorly and 

 downwards in the preoperculum two views are admissible between 

 which in the absence of further evidence it is difficult to decide. 

 Either it may as in the great majority of Ichthyopsida continue on 

 into the mandible as the operculomandibular canal or as in Chaeto- 

 stonius it may (Fig. 6 1') run into the interoperculum. 



The absence of a connection between the supraorbital system and 

 the lateral line of the body dorsally — that is the portion termed in 

 this paper "main canal" — appears very remarkable. Functionally it 

 may have been replaced by the commissure 6, 



In addition to the phenomena hitherto described in the Siluroids 

 it is worthy of note that in those forms which possess a highly de- 

 veloped dermal armour, so far as I have been able to examine them, 

 a faint line may be seen crossing the parietal bone. 

 It could not be traced at all in sections but is optically apparent on 

 careful examination of the surface. Should it indicate the existence 

 of a former canal, which is not an unreasonable supposition, it would 

 be a structure homologous with the middle dorsal line of pit organs 

 of Amia, with perhaps the transverse commissure described by Sage- 

 mehl in the Characinidae, and the transverse commissure of other 

 Placodermi e. g. Asterolepis and finally possibly the transverse com- 

 missure of (Mamydoselachus which lies in fr-ont of the openings of 

 the ductus endolymphatic (Garman). 



In Bothriocephalus, judging only from the literature to which I 

 have had unfortunately to confine myself, a canal seems to have been 

 present corresponding to an extension of branch G of the supraorbital 

 system of Clarias. 



The last groove which remains for comparison in Coccosteus is 

 the one present in the coracoid plate (V dermoclavicle) and the interest 

 which it possesses lies in the fact that the strongly marked 

 branch leading to pores 7 in Clarias and Auchenaspis, 



