[29] THE EVOLUTION OF THE FINS OF FISHES. 1009 



to crowd the actinophoral elements together to form the basipterygium. 

 This gradual crowding together of the hypiiral elements and the proxi- 

 mal ends of the rays is well shown in seven figures of tbe young o^Atheri- 

 nichthys 7iotata, Giinth., given by A. Agassiz in T*art III of his paper 

 On the Young Stages of Osseous Fishes, Plates X and XI, where the advance 

 inward of the dermal fold just spoken of is also well shown. 



There are instances, however, in which so much of the dorsal lobe 

 l)rojects posteriorly as to really constitute what we have called an opis- 

 thure, as in Lepidosteus and Gasterosteus aciileatus, Linn., where the ter- 

 minal caudal somites, with their included metameric skeletogneous tracts, 

 are wholly absorbed, whereas in most Teleosts what corresponds to the 

 opisthure of the i^recediog species is included more or less completely 

 by the tissues which enter into the formation of the permanent tail. It 

 thus happens that most of the primitive hypural elements in most forms 

 are permitted to develop, but are much crowded together proximally so 

 as to leave the urostoyle to project, in many cases permanently, and in 

 almost all species during the very early stages, in the form of the up- 

 turned eud of the chorda, which might, in its m,embrauous condition, 

 be called the urochord. In many instances amongst Teleosts degen- 

 eration or suppression of hypural elements has occurred, but even in 

 these cases the loss in number has been small. 



Those forms, such as the Salmonoids, which have the upturned end of 

 the caudal vertebral axis more or less extensively segmented (PI. VI, 

 Fig. 2), are apt to develop rudimentary epural spines, which are homo- 

 dynamous with the more anterior normal neural spines, as has been 

 conclusively shown by Lotz. Distinct in their cartilaginous condition, 

 they are finally covered with perichondrium, which becomes confluent, 

 and ossifies to become the lateral " grosse Deckstiicke" of Lotz. Such 

 epural rudiments consequently become displaced more or less upwardly 

 and laterally, apparently for the reason that the urochord is bent up- 

 wards so as to be shoved in between their homonomous basal halves. 



In those forms which do not have the axis distinctly segmented be- 

 yond the bend of the urostyle it would seem that no epural spines are 

 developed even as cartilaginous rudin^ents. J. P. McMunich's criti- 

 cism of Lotz, in his Osteology of Aniiurus (Proo. Canadian lust., Vol. II, 

 No. 3, p. 298), as based on the latter form, is therefore entirely gratui- 

 tous, as there are never any epural cartilages developed on the urostyle 

 of Amiurus, as in Salmo, as may be learned by examining Fig. 1 on PL 

 IV, taken from the completely chondritied caudal skeleton of an embryo 

 of Amiurus albidus fifteen days old. 



What is true of the pectorals tbe writer has found to be true of the 

 ventral fins of Teleosts in a number of widely-separated genera. In 

 Amiurus the material of which the axial portion of the pectoral is 

 formed consists of mesoblast continuous with the spongy connective 

 tissue of the larvje. In like manner the medullary substance which 

 fills the epiblastic fold, from which the anterior dorsal is developed \\\ 

 u Mis. 68 64 



