ANATOMY OF MEGAPTERA LONGIMANA. 25 



venter in B. musculus has, apart from what is given by the 

 rise towards the glenoid margin, very little concavity. Traced 

 upwards from the neck, it is at first concave, then convex. 

 That convexity is very strongly marked in my 64-feet-long 

 specimen, so as to give a second concavity towards the top. 

 Viewed longitudinally, the venter is, on the whole, convex, 

 owing to the bending outwards of the angles, but, for about 

 the middle half or more, between the anterior and posterior 

 beams, there is some general concavity, about \ inch deep, with 

 depths of about 1 inch at the fossae between the ridges. A 

 general concavity of the venter is better marked in B. borealis, 

 to a depth of 1^ inch, and it has the same radiating ridges as 

 B. musculus, though they are less prominent. In Megaptera 

 the venter, traced vertically, is, above where it is influenced by 

 the neck, almost flat except along the middle part where the 

 bend outwards and broadening at the top causes a slight ver- 

 tical concavity. Traced longitudinally, the posterior beam, 

 occupying the posterior half of the bone, shows a shallow con- 

 cavity along its middle following the great convexity here on 

 the dorsum. Another shallow concavity runs up between the 

 two beams, at about the junction of the anterior and middle 

 thirds of the bone, corresponding to the thin part of the upper 

 border and to part of the great concavity on the dorsum. But 

 the whole ventral surface strikes the eye as flat and smooth 

 compared with that of B. musculus. Taken from the glenoid 

 margin, the venter has a depth, in Megaptera, of 2 inches ; in 

 the 50-feet-long B. musculus, J inch less ; in the 60J-feet-long 

 one, J inch more than in Megaptera. The absence of the 

 radiating ridges for the intermuscular septa of the subscapu- 

 laris muscle, would seem to indicate a less development of that 

 muscle in Megaptera. The greater thickness of the bone has as 

 it were filled up the radiating fossaB between the ridges, but the 

 ridges were no less necessary for the fibrous septa which so 

 much increase the origin of the muscle. 



The dorsal surface in Megaptera is more simple. Traced 

 longitudinally, it is convex at each beam, the posterior convexity 

 occupying about half, the anterior about a fourth, of the whole 

 surface, with a marked concavity between, corresponding to the 

 thin part of the upper border. A line between the most 



