56 



axonosts. As therefore the first axonost of the anal fin 

 supports two baseosts completely and another partially, it 

 follows that it has been formed of at least 3 axonosts fused 

 together ; but, as before remarked, we have no direct 

 evidence of this.f The first baseost is situated in a de- 

 pression and the succeeding two on an elevation on the 

 postero-ventral surface of the first axonost, whilst the 

 second axonost is closely opposed for its whole length to 

 the same surface. The axonosts 3 to 7 fit by their tips 

 into a posterior longitudinal furrow borne on the first. 

 It is obvious that, as the first axonost is situated between 

 vertebrae 18 and 14 and axonost 7 lies in front of the 

 haemal spine of the latter vertebra, axonosts 1 to 7 and fin 

 rays 1 to 8 are situated in a space bounded morphologi- 

 cally by two adjacent vertebrae. As previously men- 

 tioned, in only one other part of the body are as many as 

 4 axonosts found in a corresponding position. 



The posterior extremity of the anal fin presents no 

 features of special interest except that the last fin ray 

 articulates with the last axonost without the intervention 

 of a baseost. 



The mechanism of the fin ray has now to be described 

 (see fig. 16). Each ray consists of two longitudinal 

 distally segmented pieces {F.R. a, h) held together by a 

 transverse ligament {F.R. e). Proximally these two 

 pieces diverge and embrace the baseost (Bs.), and also to 

 a limited extent the axonost (^4.^;.). The articular surfaces 

 bear pads of a peculiar kind of soft cartilage (M.C.). Each 

 half of the fin ray is connected with the baseost by a stout 

 ligament {F.R. d). Now the only connection between the 



f Whatever doubts may arise on this point will be settled by a reference 

 to the condition in Solea, as described by Cunningham ; and in R)w7)ibus, 

 as described by Kyle. 



