Regan^-'--' describes these as one pair of premaxillaries 

 and one pair of nasals „ but as has been shown^ the nasal bones 

 in the broadbill swordfish are very small and so this must be 

 thought a misconception^ The determination of the homology of 

 these four sets of bones must wait upon future studieSc The 

 lower jaw is made up of the articular and dentary bones,, The 

 dentary bones are stout in the Istiophoridae^ but in the 

 Xiphiidae they are rather weak with their anterior parts 

 gradually tapered. In the Istiophoridae the predentary is at 

 the tip of the lower jaw^ and the dentaries are partly joined 

 to each other and partly joined to the predentaris So In the 

 Xiphiidae there are no predentarieso On the palatines the 

 istiophorids have fine setiform teeth„ but the xiphiids have 

 no teeth on these bonesc A number of differences can also be 

 seen between the two families in the form of the various bones 

 of the opercleo There are no gill=rakers on the gill arches o 

 In the general view, the joining and articulation of the bones 

 of the skull are weaker m the Xiphiidae,. 



Figure 14,, —Structure of vertebrae, ao Sailfishj b^ Broadbillj 

 Co Short-nosed spearfisho 



Figure 15o — -Articulation of vertebrae (short-nosed spearfish)o 



The individual vertebrae in the Istiophoridae are long 

 and the central portion of the ventral side is markedly con- 

 strictedo The neural and hemal spines generally form thin and 

 broad flat plates,- On the two most anterior vertebrae,, however, 

 the neural spines are more or less spine^shapedo These lamellate 

 neural spines are surrounded by the neural processes of the 

 following vertebrae so that they are firmly stuck together. The 

 neural processes have been modified into broad lamellar bodies 

 projecting far forward,. Their anterior edges extend almost to 

 the middle of the preceding vertebrae,. The hemal spines on the 

 first two vertebrae are narrower than the neural spines of the 

 same vertebrae^ but on the more posterior ones they are of 

 roughly the same shape and sizeo The hemal processes are narrower 

 than the neural processes and are somewhat rod-shapedo They pro- 

 ject far forward and encompass the hemal processes of the verte- 

 brae immediately preceding themo The relative length and height 

 of the neural and hemal spines differ considerably from one 

 species to another. The neural canal and the hemal canal are 

 both quite large,, There are no transverse processes and the ribs 

 are without heads /7'?_7"^ The vertebral count is 24 for the 

 whole family J, but some are shown by the formula 12 / 12 e 24 

 and others by the formula 11 ;^ 13 s 24c The former is the type 



(^-^ Regan, C, T„ Ann„ Mag, Nato Hist„ , London^ Vole 3^ Noc 13, 

 pp, 73c 



36 



