2. Skeleton 



Tziking a general view of the skeleton, it differs conspicuously from the scombroid 

 fishes, which are supposed to be most closely related to these species. The differences between 

 the fishes of the families Istiophoridae and Xiphiidae are also remarkable, and it must be said 

 that it is very difficult to compare and describe the skeletons of the fishes of these two families 

 at the same time. Consequently I shall concentrate here on setting forth the comparisons among 

 the fishes of the family Istiophoridae and will not touch in any detail upon the broadbill except 

 in special circumstances. 



(a) Craniuna 



There are no striking differences in the cranium. However, some differences are seen 

 in the proportion of its length to its breadth. If the crania of these fishes are compared with 

 those of the scombroids, striking differences can be seen. Most conspicuous, of course, are 

 those differences based on the construction of the snout, which is a peculiarity of these fishes, 

 but the thin processes on the supraoccipital which are found in the scombroids do not appear in 

 these fishes, and the posterior projections of the epiotic and pterotic are conspicuous. In all 

 of the species there are marked differences among individuals, so that it is difficult to deternnine 

 the basic type (plate 3). 



(b) Vertebrae 



The vertebrae of the istiophorid fishes are very characteristic. The vertebrae of the 

 fishes of this family and of the scombroids differ markedly both in form and in nunnber. In the 

 scombroids the number of vertebrae is fairly great, being from 38 to 50j_L'. In the istiophorids 

 in all genera there are 24 vertebrae and both the neural spines and hemal spines are flattened into 

 broad plates. 



These flat plate-like neural spines are inserted into the neural processes of the succeeding 

 vertebra so that they are strongly articulated together. The neural processes are modified into 

 broad plate-like structures which extend far forward, so that the tips reach almost to the nniddle 

 of the next most anterior vertebra. However, in the two nnost anterior vertebrae the neural 

 spines are far narrower than the more posterior ones and are shaped more nearly like spines. 

 The hemal spines are roughly the same shape as the neural spines and are strongly fused to the 

 hennal processes of the neighboring vertebrae, just as are the neural spines and neural processes. 

 Both the neural and hemal canals are very large (plates 4 and 5). 



It is inferred from their form that the vertebrae of the istiophorid fishes are most 

 unsuitable for precise movements and that sudden changes of direction are probably impossible. 

 It is a well known fact that fishing boats are often attacked by spearfishes, but it is doubtful 

 whether such attacks are nnade purposely. It may be that the fish strikes the boat because the 

 construction of the vertebrae does not give it the ability to change its course in tinne. 



These fishes of the fannily Istiophoridae, with 24 vertebrae, can be further divided into 

 different groups. In the first group the vertebral count can be shown by the fornnula 12 + 12 and 

 in the second group by the formula 11 + 13. Considerable differences in the form of the vertebrae 

 can, of course, be detected among the species (plate 4), the proportion between the length and the 

 height of the vertebrae differing in each species. The hemal spine of the first caudal vertebra 

 differs markedly in form from those of the more posterior vertebrae and also differs among the 

 species (see plate 5). 



The above-mentioned two groups based on the vertebral counts have no connection at all 

 with the presently recognized classification of the istiophorid fishes. The first group includes the 

 shortnosed spearfish ( Tetrapturus angustirostris ), the sailfish ( Istiophorus orientalis ), and the 



— Kishinouye, K. , Jour. Coll. Agri. Tokyo Imp. Univ., Vol. 8, No. 3, 1923. 



11 



