306 Mr. R. I. Pocock on the EquiJai. 



Apart from the retention of the stripes and the unique 

 character of the reversal of the spinal hairs, the mountain 

 zebras seem to diflfer but little from the African wild asses. 

 In general form, shape of head, length of ear, and narrowness 

 of hoof the two species are strikingly alike. As contrasted 

 with what obtains in the following group, tliere is a marked 

 tendency of the stripes to persist upon the legs in the species 

 that have lost them on the body. 



{E. qungga), 



The validity of this group must, I think, be admitted. 

 The true quagga is generally set apart as an isolated species, 

 displaying less similarity to Burchell's zebra than the latter 

 shows to the mountain zebra. Judging, however, from 

 extant figures and the few specimens of E. quagga preserved 

 in museums, considerable individual variation ^Yas exhibited 

 in the extent to which the stripes on the flanks and hind- 

 quarters were suppressed. Even in the extremest cases of 

 suppression the difference between E, quagga and the typical 

 E. Burchelli is less pronounced than tliat between the latter 

 and the subspecies known as Craicshayi ov Granti'*. 



The quaggas furnish an admirable object-lesson in sub*- 

 specific forms. Broadly speaking, the most fully striped 

 forms occur to the north of the Zambesi, both Grant's quagga, 

 which inhabits Southern Abyssinia and British East Africa, 

 and Crawshay's quagga, from British Central Africa (Nyasa- 

 land) , being strongly striped to the hoofs and banded to the 

 middle line on the belly. Nearest to these comes Selous's 

 quagga, irom Mashonaland, which, however, has paler narrow 

 intermediate stripes, at least on the hind- quarters. From this 



/ The adoption of the term " quagga " for this group will be followed 

 by the advantage of enabling us to speak of these subspecies as Crawshay's 

 or Grant's quagga, instead of Crawshny's Burchell's zebra, Grant's Bur- 

 chell's zebra, &c. 



