1904.] MALAGASY BAT MYZOPODA AURITA. 5 



temminckii in its proportions and the angle at which it is set. 

 It is even less expanded across the ilia, which are rounded instead 

 of being concave above. Its length, dorsally, is 9 mm. ; its breadth 

 at the acetabulum 5*4, and in front across the ilia 3'8 ; its depth 

 from front to back 4" 6. The symphysis is more slender, though 

 still firmly ossified, and the obturator foramen is smaller. The 

 pectineal process is unusually slightly developed, far less than in 

 any Bat that I have had the opportunity of examining, and is 

 indeed hardly to be called a process. 



In its entirety, therefore, the skeleton of Myzopoda is remark- 

 able for the simplicity and non-specialisation of the diflTerent 

 parts. In all other Bats' skeletons that I have examined (not, I 

 confess, a very large number, but still representing most of the 

 grovips) there are some specialisations, such as unusual ankyloses or 

 peculiar development and expansions of processes, but in Myzopoda 

 none of these are present, while even normal Ohiropterous 

 charactei's, such as the elongation of the coracoid and the pectineal 

 process on the pelvis, are at a minimum. 



It will be evident fi^om the above description, from those of 

 Milne-Edwards, Gi'andidier, and Dobson, and from the figures 

 now published, that Myzopoda is a most remarkable and peculiar 

 kind of Bat, and that it must certainly form a s]3ecial Family, 

 the Myzopodidce, though the affinities of this Family are by no 

 means clear. Both Milne-Edwards and Dobson referred Myzopoda 

 to the Yespertilionidie, forming of it a group equivalent in value 

 to the Plecoti, Vespertiliones, and Miniopteri. Now, however, that 

 Mr. Miller has taken JSfatalus and Thyroptera out of the Vespei-- 

 tilionidse *, the chief reason for leaving it in that Family has been 

 removed, for it is only to those genera that Myzopoda could be 

 supposed to be related. 



On the whole, it appears to me probable that Myzofoda is most 

 nearly related to the Natalidse and Mormoopidee t, being a 

 descendant — specialised in some characters and primitive in 

 others — of the common ancestors of those groups. The occur- 

 rence of sucking-disks both in Myzopoda and Thyroptera is no 

 doubt a coincidence, for these organs would appear to be com- 

 paratively recent specialisations. On the other hand, the 

 possession of three phalanges in the middle finger in the same 

 two genera is presumably a primitive character, for both Phyllo- 

 stomatidse and Mormoopidse possess three, and the loss of the third 

 one in Natalus and Furipterus of the Natalidse % is one of the 

 characters in which that Family shows an approach to the still 

 more highly-developed Yespertilionidse. 



* '•' Historj' and Characters of the Family Natalidee," Am. Nat. xii. p. 245 (1899). 



f As has already been done by various other authors, I would regard the 

 " LohostomincB " of Dobson as forming a Family distinct from the Phyllostomatidae. 



X Mr. Miller states that the long terminal phalanx of the two-phalanged Natahis 

 and Fwipterus " is divided into two in Thyroptera" but without very definite 

 embryological evidence it would be difficult to accept this horaologisation. The 

 tei'minal bar of cartilage beyond the second phalanx in the Vespertilionidae is a 

 hint at a different homology, and one more in accordance with the usual ideas on 

 the subject. 



