234 On ihe Affinities of the Genus Lipotes. 



Two pairs of subcutaneous air-sacs are sent off from each 

 side of the spiracle — an anterior pair dorsal to the floating 

 bones, and a posterior pair veiitial to those structures. Tiiere 

 is n)aikcd asymmetry between the two sides in this region. 

 On the left side the air-sacs and the floating bone are very 

 materially smaller than on the right, while the posterior air- 

 sac is wholly concealed beneath the inner portion of the bone. 

 On the right side the posterior air-sac attains an enormous 

 development; passing out from beneatii the edge of the bone, 

 it covers the whole surface of the dilator naris, and in front 

 it sends a diverticulum upwards over the dorsal surface of the 

 bone, the end of the diverticulum actually abutting broadly 

 against the outer wall of the anterior air-pouch. This dorsal 

 diverticulum, although but a small part of the right posterior 

 air-sac, has a much greater capacity than the whole pouch of 

 the left side. In the respects described the right side appears 

 to be much more highly modified than the left, so that we 

 may sa}', ])erhaps, that Lipotes affords us two distinct stages 

 of evolution simultaneously. 



The floating bones, above described, may represent an early 

 stage in the development of a bony facial mask, tlie extra- 

 ordinary facial structure of Platanista then representing the 

 culmination of such a process. By possessing even rudiments 

 of such a structure Lipotes would be well on the way towards 

 Platanista, and brought into closer relation with the latter 

 genus than with any other. All the characters in which 

 Lipotes resembles the South-American Inia are, perhaps, 

 primitive features common probably to all the more primitive 

 membeis of tiiis group, and they, in all probability, were 

 shared by the ancestors of Platanista. 



Tlie stomach, too, when compared with that of Platanista^ 

 is of very j)rimitive form, tlie veiitriculus being widely 

 confluent with what re[)resents the second compartment in 

 Platanista and other dolphins. The stomach, therefore, may 

 be described as being less completely segmented proximally 

 than in most other genera, although towards the pylorus 

 several small compartments are shut off as usual. 



To sum up, we are inclined to believe that Lipotes is more 

 closely related to Platanista than to any other known genus — 

 a conclusion in liarmony with its distribution, — and that it 

 lepresentsin many respects an early stage in the evolutionary 

 |jrocesses which have led to the development of Platanista. 



