NO. 1&76. TREE8HREWS: FAMILY TUPAIID^—LYON. 9 



that seen in the squirrels; and so in every respect is the incus. All 

 the ossicula differ from the varied forms in other families of the 

 insectivora, except that the incus somewhat approaches in type that 

 of Galeopithecus. Taken as a whole, the ear bones of Tupaia are 

 higher in type than in any other family belonging to this order."^ 



Vertebral column — Five skeletons in the United States National 

 Museum have the vertebral formulae: 



Cat. No. 124317, Twpaia glis fenuginea, C, 7; D., 13; L., 6; S., 3; C, 24. 

 Cat. No. 49468, Twpaia lacernata vAlhinsGni, C, 7; D., 13; L., 6; S., 3; C, 27. 

 Cat. No. 174609, Tupaia demissa, C, 7; D., 13; L., 6; S., 3; C, — . 

 Cat. No. 111782, Tupaia nicobarica, C, 7; D., 13; L., 6; S., 3; C, 28. 

 Cat. No. 174611, Tana tana, C, 7; D., 13; L., 6; S., 3; C, 22. 



Certain authors give the lumbar series as 5, 6, or 7.^ Blainville ^ 

 in the text of Osteography of Mammals says that the number of 

 lumbar vertebrae is seven, but on studying the first plate I can count 

 only six lumbars. By the same author ^ the sacral vertebrae are 

 said to be two, and the caudal 22-23, figures which are confirmed 

 by the plate. 



The vertebral formula of Ptilocercus is given by Jentink* as D. 14, 

 L. 5, S. 3, Cd. 31. 



The individual vertebra3 are well developed, and with well-defined 

 processes, that is relatively better developed and with better developed 

 processes than in the case of the Macroscelididge. The first six 

 cervical vertebree are pierced by vertebral foramina; the atlas is 

 large and relatively heavy, and so is the axis which bears a conspicu- 

 ous posteriorly directed dorsal spine. On all the other cervical ver- 

 tebrae dorsal spines are absent or represented by minute projections. 

 As is usual, the under and lateral surface of the sixth cervical is well 

 developed with a prominent costal or pleurapophysial plate. The 

 dorsal vertebrae show no noteworth}^ peculiarities. The dorsal proc- 

 ess on them is rather low and directed backward till the ninth is 

 reached, which has the dorsal process directed upward; the dorsal 

 processes on the three remaming dorsal vertebree are low and directed 

 forward. The lumbar vertebrae taken as a whole have well-developed 

 anteriorly directed dorsal processes, and still better developed, wide 

 anteriorly directed transverse processes, the processes being smaller 

 on the anterior vertebrae of the series. In Alacroscelididae, repre- 

 sented by a skeleton ElepTiantulus roseti, the transverse lumbar 

 processes are shorter and wider. The sacrum m Tupaia and Tana is 

 composed of three firmly fused vertebrae, that is, no large foramina 

 exist between the transverse processes of the first and second, and 

 second and third sacrals, as is the case in Macroscelididas. The first 



1 Doran, Trans. Linn. Soc. London, ser. 2, 1S79, vol. 1, p. 444. 

 > See Gregory, Bull. Amer. Mus. Nat. Hist., vol. 27, 1910, p. 275. 

 s Ost6ographie des Mamniiferes, vol. 1, p. 33, pi. 3, lower figure. 

 * Notes, Leyden Museum, vol. 7, 1885, p. 38. 



