CLASSIFICATION OF THE SCIURID^. 219 



middle below and above in its distal third. The blade, about one- 

 third of the length of the entire bone, is lodged in the distal con- 

 cavity. It has a straight upper edge, not rising above the median 

 dorsal convexity of the shaft. Its posterior third is pointed and 

 free from the shaft. Its anterior end is attached immediately 

 behind the slightly expanded tip of the shaft. Hence the blade 

 is at the distal end of the shaft. The baculum above described 

 measures 26 mm., the longest in the Sciuridse. (Text-figs. 18, 

 E, F; 21, B.) 



CaUosciurus notatus, from Java, has the glans shorter than in 

 C. prevosti and with a pronounced dorsal curvature in its distal 

 half, so that its tipper surface is concave, its lower surface convex. 

 The sharp distal edge formed by the blade is more pronounced, 

 and the orifice is situated as in C prevosti. The baculum difi'ers 

 from that of C. prevosti in being relatively stouter, in having 

 a larger thickening at the proximal end, a strongly upcurled 

 distal half with a concave upper and a convex lower edge ; but 

 the blade, although similarly placed distally, is not so pointed 

 and its upper edge is above the line of the edge of the shaft 

 preceding it. (Text-figs. 18, G, H ; 21, C.) 



In specimens of C. castccjieoventris from Ningpo and of 0. airo- 

 dorscdis from Siam the shaft of the baculum resembles that of 

 C. notatus, although it is less strongly curved upwards ; but the 

 blade, which is considerably longer, being nearly half the length 

 of the shaft, is not attached to its distal end, which projects freely 

 beyond the anterior end of the blade, the blade itself lying nearly 

 in the middle of the dorsal concavity of the shaft. In C. casianeo- 

 veniris the baculum measures 19 mm. (Text-fig. 21, A.) 



In Thomas's pi-eliminary account of the bacula of the Oriental 

 Squirrels, (7. prevosti was one of the species in the list of un- 

 examined forms, whereas C. notatus, castaneoventris, and atro- 

 dorsalis appear in the list of species of which the baculum was 

 examined. The remaining species of this list were caniceps, 

 erythrceus, pluto, and sladeni. 



The differences between the bacula of C. prevosti and C. notattcs 

 on the one hand and of C. castaneoventris and C. atrodorsalis on the 

 other suggest the possibility of usefull}'' splitting Calliosciurus 

 into two genera. The characters appear to me to be at least as 

 valuable as the presence or absence of a degenerate premolar. 

 But it would perhaps be premature to take that course, at all 

 events until it has been shown that the bacula of other species do 

 not show intermediate stages in the position of the blade. 



The baculum of a specimen of Alenetes herd'morei from Kokareet 

 (Tenasserim) in the British Museum very closely resembles that 

 of G. prevosti described above, except that the shaft is straight 

 up to the concavity containing the blade. It is convex below 

 only at its distal end and correspondingly concave above, the 

 blade lying in the distal concavity as in (J . pirevosti and C. notatus. 

 The baculum of this species, measuring 11 mm., was held by 

 Thomas to have a blade somewhat connecting the types of bacula 



