64 PROCEEDINGS OF THE NATIONAL 3IU8EUM. vol. 45. 



Diagnosis. — Tupaia chinensis is characterized by a unifrom 

 grizzled olivaceous gray color without ferruginous on upper parts, or 

 ochraceous colors on rump; skull slightly smaller than in T. glis 

 ferruginea, with relatively shorter rostrum; mammas, 3-3=6. 



Color. — Upper parts of head, neck, and body and outer side of legs 

 a fine uniform grizzle of blackish and a color that varies from buff to 

 ochraceous in certain individuals, the lighter colors predominating 

 anteriorly; both colors are about equally mixed, in some individuals 

 the lighter colors are in excess, and in others, especially in the middle 

 line posteriorly the black admixture is sometimes in excess ; the tail 

 above is a coarse grizzle, sometimes showing indistinct annulations 

 of the same colors as have the upper parts of body, below the tail 

 is lighter, especially in the middle line; the underparts of head, neck, 

 and body, including the inner sides of the legs, vary from distinct 

 whitish, sometimes with the dark bases of the hairs showing through, 

 to buffy; hands and feet similar to outer sides of legs but often lighter 

 and grayer; shoulder stripe poorly developed and sometimes prac- 

 tically obsolete. Three skins from Meechee, China, are quite light 

 and grayish, but are almost exactly matched by a British Museum 

 skin from Manipur, Reg. No. 85.8.1.89. The Darjiling and two of the 

 Tura skins are rather dark, as are also the Siamese skins, none, 

 however, are so generally dark as are Tupaia concolor and T. modesta. 



Slndl and teeth. — These are of the same general style as they are in 

 T. glis ferruginea, but average slightly smaller and have a relatively 

 shorter rostrum, so that the distance from the end of the premaxillary 

 to the lachrymal notch is generally distinctly less than the distance 

 from the notch to the external auditory meatus. Although T. 

 chinensis was originally separated from T. hclangeri mainly on skull 

 characters, I have been able to find no satisfactory constant char- 

 acters to distinguish skulls of the two forms. The individual varia- 

 tion in skulls of Tupaia is quite considerable, and with a relatively 

 small number of specimens such as Anderson seems to have had it 

 would be comparatively easy to find distinguishing features. I 

 regret that I have not seen his type or cotypes. (Plate 8, fig. 1.) 



Measurements. — Anderson's measurements of the cotypes con- 

 verted to millimeters: Head and body, , 165 mm.; tail, , 



156; hind foot, , 40; inferior margin of foramen magnum to tip 



of premaxillaries, 39, 40 (making a condylobasal length of approxi- 

 mately 42, 43) ; zygomatic width, 22, 23. The usual measurements of 

 adults corresponds very closely with those of the cotypes, the head 

 and body measurement is often 5 mm. longer, but the tail in most of 

 the specimens which have collectors' measurements is nearly always 

 from 5 to 15 mm. longer than head and body, but the skins and 

 alcoholics of which I have taken approximate head and body, and tail 

 measurements show the tail to be shorter than head and body. The 

 maxillary tooth row is about 17.5. For details of measurements, see 

 table, page 66. 



