1 68 ALUS. [Vol. XVIII. 



ward and backward to the ventral edge of the mass, where they 

 each end in a single pore. Either the seventh or eighth system 

 was wanting in the specimen shown in Figure 3. 



Between the eighth system and the lower edge of the post- 

 frontal bone there is no dendritic system. The innervation, how- 

 ever, shows that two systems at least, Nos. 9 and 10 of the line, 

 should have been normally developed in this part of the canal, 

 and, in one specimen, a short closed tube was found representing 

 one of them. Both systems have doubtless disappeared as a re- 

 sult of the closing of their surface openings by the gelatine-like 

 mass that surrounds the eye, the tubes thus having been rendered 

 functionless. The associated nerves and organs were, however, 

 retained. One of these nerves enters the second bone of the post- 

 orbital series, counting from below, the other the third, those two 

 bones also lodging the related organs. In the small and delicate 

 scale-like bones between the third postorbital bone and the post- 

 frontal no indication of a nerve, organ, or dendritic system was 

 found in any specimen. 



System No. 11, the next system in the line, was represented, in 

 all the specimens examined, by three tubes and pores. Two of 

 these three tubes always left the main canal as it traversed the 

 postfrontal and the dermal tissues dorsal to it, the other one aris- 

 ing at the point where the canal enters the open groove between 

 the edges of the frontal and squamosal. The first two tubes are 

 directed backward and slightly upward, one lying partly in the 

 postfrontal, and the other wholly in the dermal tissues superficial 

 to that bone. The third tube runs laterally and downward from 

 the orbital surface of the canal, at the point where it is joined by 

 the supraorbital canal. The external openings of all three tubes 

 lie slightly dorsal to the gelatine-like mass around the eye. Al- 

 though the development of this system was not traced from 

 younger stages upward, it is sufficiently evident that it is a double 

 system formed bv the fusion of svstem 11 infraorbital with system 

 7 supraorbital, and that it corresponds to the double system 15 

 infraorbital — 7 supraorbital of Aniia. In the subdivision of this 

 double system, in Aniia, the separation of the two parts took place 

 along the squamosal part of the line ; in Scomber it takes place 

 along the suborbital part. A variation in the tubes and canals, 



I 



