NO. 9 MAMMALS FROM CAVES IN HAITI — MILLER I3 



BROTOMYS (?) CONTRACTUS sp. nov. 

 Plate 2, fig. 2 



Type. — Anterior portion of skull, lacking zygomata, nasals and 

 teeth, No. 253100, U. S. Nat. Mus. Collected in small cave near 

 St. Michel, Haiti, March, 1925, by Gerrit S. Miller, Jr. 



Characters. — Resembling Brofomys voratus, but size slightly less, 

 rostrum relatively shorter, interorbital region narrower in proportion 

 to f rontopalatal depth and more arched transversely ; teeth broader 

 than in Brofomys %'oratus, and palate noticeably constricted, its inter- 

 alveolar width at middle conspicuously less than transverse diameter 

 of the adjoining alveoli. 



Skull. — While reseml)ling in a general way that of Brofomys 

 vorafus the skull of B. ( ?) confracfus. even in the imperfect condi- 

 tion of the only known specimen, shows well marked differential 

 characters. Most conspicuous among these is the narrowness of the 

 bony palate as compared with the very wide alveoli of the anterior 

 cheekteeth. In three specimens of B. vorafus (the type from the 

 Dominican Republic and two from Haiti) the w^dth of the palate 

 between the alveoli of the second cheekteeth is 2.55, 3.0 and 3.0. 

 respectively, and the width of the first alveolus is 2.25, 2.25 and 2.30. 

 In the type of 5. (?) confracfus the width of the palate at the same 

 level is only 1.65, while that of the first alveolus is 3.60. The narrowing 

 of the skull shown by the palate is also evident when the interor- 

 bital breadth is compared with the fronto-palatal depth. In the type 

 of Brofomys ( ?) confracfus the ratio of this breadth (15.6) to depth 

 (13.0) is only 83.3, while in the three specimens of B. vorafus it is 

 92.5, 90.0 and 92.6. The greater transverse convexity of the interor- 

 bital roof is a character which cannot be expressed by measurements ; 

 it is immediately obvious when specimens are compared in posterior 

 view. Because of the imperfect condition of the skull a comparison 

 of the length of the rostrum with anything but the length of the 

 palate is difficult; hence the apparent shortening of the rostrum may 

 be due in part to an actual lengthening of the palate to accommodate 

 the enlarged teeth. In Brofomys ( ?) confracfus the length of the 

 palate (9.4) measured from posterior border to level of anterior 

 margin of alveolus of pm'^ is essentially equal to the distance fron] 

 the latter level to alveolus of incisor (9.8) ; in B. ( ?) confracfus it is 

 barely more than the distance from alveolar level to front of incisive 

 foramina (that is, about 5 mm. less than the distance to alveolus of 

 incisor). The alveolar length of the toothrow in the type oi B. (?) 

 confracfus cannot be exactly measured (the alveolus of 7;:^ is entirely 



