846 Mi{. u. I. I'ococK ON THE [June 14, 



wisely and advantageously sejiarated t'loni the African Trngela- 

 plunes. Possibly a special subfainily slioidd be erected for 

 Aininodorcas, which at present I am unal)le to classify ; and a 

 good case could be made out for splitting up the Rupicaprinae, 

 which no one can claim to be a homogeneous unit. But our 

 knowledge of the mutual affinities of the various sections of the 

 Bovidse is so surprisingly incomi)lete that no attempt at a final 

 classification of them can at present be attem^jted. 



Owing to the impossibility of arranging the subfamilies 

 according to their probalile reliitionsliijts ii\ a linear series, it 

 comes about tliat in the annexed t:i]ile divisions of this rank ai'e 

 juxtaposed which have obviously no near relationship to each 

 otliei'. OreotrcKjn.s, for instance, is jirobably, as is usually supposed, 

 allied to tlie Xcoti'aginjw and Pantliolops to the Saigina% Antilo- 

 pina', or Caprin:e. Yet these two genera fall, by the diffeiential 

 characteis made use of, under the heading f alongsitle the 

 ^-Epycei'ina^ the Orygina', and others with which they have 

 certainly, as I at jjresent believe, no close connection. That is 

 one of the obvious defects of such tables as the subjoined, which 

 are commonly much more useful than scientific. 



I have made no attempt to include Orihos in this key, Lcinn- 

 berg's recent paper on this form making one conclusion abiuidantly 

 cleai', namely, that the genus must for the present be left by itself 

 in the subfamily Ovibovina». 



a. On till' t'vont and usually also on the hind li'i;- a lai';j;c inter- 



di;;ital clct't, forming a ^-lantl the oritifr of which is a lonir 



slit with its u])]ier niary;in sharply (letint'il hy the abrujjt 



desi-ont of the intcguinont hotwecn the digits ; or else the 



orifice of the cleft is sliorter and the gland more ov less 



tubuliir, extending up between the hones of the pastern, its 



posterior wall heing formed by ii close fold of integuinent 



forming the heel-tie inferiorly, so that the orifice of the 



gland opens downwards between the hoofs. 



h. Preorbitiil gland nearly halfwaj' between the eye and the 



nose, not valvular or invaginated, hut marked externally 



by a l)are streak of skin studded serially with pores; web 



of ])astern reaching only to the heels, closely folded, the 



gland tubular with the ovitice opening freely into the 



space biitween the hoofs CEpriAi.oPHTN.i-; {p. 8t>7). 



b'. I'reorbital gland occasionally aborted, when i)resent usually 



a valvular invagination, never'uiarked externally liy a 



streak of integument with serially /irranged pores. 



c. (ilaiids eiiually or a]iproximately c(inally ileveloped on 



both front and hind feet. 



d. M utile large, naked above as fur buck as the ))osterior 



angle of the nostrils Neotbaginjj* (j). 879). 



d' . Muffle snuill, hairy above almost or (piite to the 

 anterior angle of the nostrils. 

 e. IV'dal glands mure or less tubular; the integument 

 of the back of thcpusteni closely folded at the heels. 



M.A.i)0(}i'iN-i'; (p. 876). 



• Hased upon observation of Oiirihla and Riijihirrrits, the other genera being 

 unknown to ine. 



