354 Mr. J. W. Davis on the Teleostean 



extending '6 of an inch towards the base ; the anterior sur- 

 face is produced so as to form a median keel, which extends 

 all the length of the spine ; towards the base this keel is cor- 

 rugated, forming a line of rough granulations. The sides of 

 the spine are slightly striated. All the spines have an internal 

 pulp-cavity, but show no trace of the posterior hollow seen 

 in the Elasmobranchs. 



In numerous species there is a dorsal- fin spine, but no 

 spines to the pectorals. 



In a small specimen of Macrones from one of the Indian 

 rivers there are two pectoral and one dorsal fins which have 

 spines attached. They present a close resemblance to some 

 of the spines of the fossil Comps acanthus. The pectoral fins 

 are armed with a strong spine "5 inch in length. The spine 

 tapers to a point. Its posterior face is straight, the anterior 

 very slightly curved. It is somewhat flattened laterally, and 

 ornamented with longitudinal strise. On the median posterior 

 surface is a row of denticles, which extend along the whole 

 length of the spine exposed ; the denticles are equal in length 

 to the breadth of the spine, sharply pointed and slightly re- 

 curved towards the base. Attached to the spine is the pectoral 

 fin, composed of seven or eight bifurcating fin-rays. The 

 dorsal spine is about two thirds the size of the pectoral ; it is 

 straight, pointed, striated longitudinally, and has on its pos- 

 terior surface a single row of short straight denticles, which 

 point towards the base of the spine at an angle of 45°. It 

 is not connected with the dorsal fin, but stands detached. 

 It is implanted in a triangular bone formed by a pro- 

 longation of the bones composing the occipital region of 

 the head. 



As already explained, the spines of the Siluroids are at- 

 tached by an articulation of greater or less complexity ; but I 

 must confess there does not appear to be any thing extraordi- 

 nary in this difference. In Pleur acanthus the spine is adapted 

 in the best way to the cartilaginous skeleton of the fish, whilst 

 in the Siluroids, having a bony and firm base, nothing appears 

 more reasonable than that, during the period of this develop- 

 ment or change in the character of the skeleton of the fish, the 

 spine should have been adapted to its support, and have ac- 

 quired its present method of attachment. There is also the 

 important consideration that the spines of the rays are all 

 similar in character, with two rows of denticles situated on 

 the lateral surfaces of the spine as widely separated as is 

 possible ; whilst those of the Siluroids exhibit every gradation 

 from spines with the two rows of denticles placed laterally, as 



