1893.] CETACEAN- GENTTS MESOPLODON". 225 



further, and moreover it is not ossified to the other bones, and, as 

 a matter of fact, it remains for a long time separate. It may 

 become one with the mesorostral bone in very aged animals ; even 

 then the suture remains generally very distinct, wedged in between 

 the upgrowths of the mesorostral. 



A section of the same snout (Plate XV. fig. 3) taken more 

 anteriorly is also of great interest, for there the thickening and 

 ingrowth of the premaxillary bones are seen to bisect the vomer 

 into two parts just below the spout ; the growth of premaxillary 

 ossifications (pmx.o) on both sides has compressed and folded to- 

 gether in the middle the vomerine walls, thickened already by 

 proliferation of their tissue, the point of union being with some 

 care observable in the median line. 



In a still more anterior section (Plate XV. fig. 4, v) the keel 

 portion of the vomer below the bisection has increased in growth 

 and appears as a round rod, part of which shows on the palatal 

 surface, and has begun to become implicated in the ivory-hke ossifi- 

 cation which has commenced. 



In the Chatham Island specimen and in that in the Otago 

 Museum (G and F respectively in above list) much the same 

 changes occur. In some cases, as, for instance, in the female speci- 

 men (H in the above list) in the Canterbury Museum, the filling- 

 up of the vomerine spout has proceeded more symmetrically, and 

 we have then greater regularity in the form of the section of the 

 snout (Plate XIV. fig. 4, a). Fig. \b, Plate XII., represents a 

 section through the middle of the snout of the type (male) speci- 

 men of M. grai/i in the Canterbury Museum, and how widely it 

 differs from that of the female of the same species in the same 

 Museum (Plate XIV. tig. 4 a) or of 31. australis in the British 

 Museum (Plate XIII. fig. 2) is at once apparent— yet not greater 

 than the difference between the three forms of Mesoplodon layai-di 

 shown in the sections a, h, c, fig. 2, p. 228. Pig. 2, Plate XIL, is a 

 reproduction of the section of M. haasti from Sir W. Flower's paper 

 in the Trans. Zool. Soc, so often referred to, and which, as he has 

 pointed out, differs so much from the section of any other he has 

 examined that he could not include it in any known species ; while 

 fig. 1 a, Plate XII., is a section made by me of the type specimen of 

 M. grayi in the Canterbury Museum, somewhat more anterior than 

 fig. 1 6, but still in the region where the vomer appears on the 

 palatal surface, and their similarity will be at once admitted. 



I have already quoted Sir W. Flower's remark that if so great a 

 change can take place, due to individual variation, as exists between 

 M. grayi and M. haasti, then most of the fossil species based solely 

 on the form of the rostrum are quite valueless. If we take, for 

 instance, the two forms M, angulatas and M. merUlineatus, there 

 exists far less difference between them than between some of the 

 forms of M. grayi or of M. layardi. 



The median lines or sutures on the surface of the mesorostral 

 bone, which vary so much, and also the gibbosities of the pre- 

 maxillai'ies, are, after studying the sections of immature forms, 



Pkoc. Zool. Soc— 1893, No. XV. 15 



