CHAEACTEBS OF SCAETURUS AND OTHER JERBOAS. 681 



Apart from the alterations necessitated by the nomenclature 

 noAV in vogue and the addition of several recently proposed or 

 recently restored generic names to the Jerboa family, I have no 

 modifications of this classification to suggest, except the severance 

 from the rest of the Jerboas of the two genera Pygeretmus 

 {Platycer corny s) and Cardiocranius* , which agree in having broad, 

 flattened, lanceolate tails, distinguishing them at once from the 

 rest of the family. They may constitute the subfamily Pygeret- 

 minee, which appears to be a specialised ofishoot of the five- 

 toed Allactaga-gvou^. The two genera differ at least in the 

 following characters : — Pygeretmus has lost the anterior upper 

 premolai', which Cardiocranius retains ; the latter has grooved, 

 the former ungrooved upper incisors. 



Lyon divided his Dipodinse into two groups : (1) i)ipM5, with 

 its subgenera, possessing three toes, a considerably inflated 

 mastoid, the upper incisor grooved, the anterior upper premolar 

 absent, and the antorbital canal for the nerve complete. 

 (2) AUactac/a, with its subgenera, has more than three digits, the 

 mastoids not much inflated, the upper incisors without grooves, 

 the small antei-ior upper premolar present, and the antorbital 

 canal for the nerve incomplete. 



If the above-mentioned characters held good there would be 

 strong reasons for separating Allactaga and its kindred from 

 Dipus and its allies as a special subfamily Allactaginse. But 

 there seem to be too many intermediate forms to make such a 

 course advisable. In the number of the toes, for instance, 

 Scarturus is precisely midway between Dipus or Jaculus and 

 Allactaga. In the Asiatic three-toed Jerboa, Dipus t, the 

 mastoids are' much less inflated than in the African forms, 

 Jacidus and Scirtopoda ; Dipus thus connects those genera with 

 Allactaga in that particular. In Dipus, too, the anterior small 

 premolar is retained as in Allactaga, whereas it is lost in Jacuhts 

 and Scirtopoda. On the other hand, the small five-toed Asiatic 

 Jerboa, Alactagidus acontion, resembles the three-toed African 

 genera, Jacidios and Scirtopoda, in the loss of this molar. 



Of the characters mentioned by Lyon, therefore, there remain 

 as distinctive only the presence or absence of the grooves in the 



* Satunin, Ann. Mus. Zool. St. Petersb. 1902, p. 582. The genus Cardiocranius 

 has au extvaordinai-il}' wide and inflated cranial portion of the skull, which, as com- 

 pared with the facial portion, is much larger even than in EucJioreutes. I associate 

 this genus with 'Pygeretmus on the assumption that the peculiarly modified tail 

 has not been developed twice independently within the group. 



t For the sense in which this genus is here used, sagitta being its type-species, 

 see Thomas, Ann. Mag. Nat. Hist. (8) ii. p. 308 (1908). Trouessart seems to have 

 been unaware of this paper by Thomas when he published his ' Faune des Mamm. 

 d'Europe ' in 1910. At all events, he proposes the new name Dipodipiis for the 

 group of Asiatic three-toed species, with sagitta as the type, for which Thomas 

 showed the old name Dipus to be available. Satunin ^eems to have come in- 

 dependently to the same opinion on this point as Thomas (Mittli. Kankas. Mus. 

 1907, p. 72). Trouessart also resuscitated Scirtopoda in quite a different sense 

 from that in which Thomas employed it (cf. supra, p. 679). But, since Thomas 

 was the first reviser of the nomenclatui-e of these Jerboas, I follow his decisions. 



