64 JOURNAL, BOMBAY NATURAL HIST. SOCIETY, Vol. XXVI. 



white or lemon yellow colour, separated on each side from the rich 

 tint of the flanks by a sharp line of demarcation. 



In 1866, de I'lsle {Ann. 8c. Nat. (Zool.) IV, p. 173) described a 

 series of breeding experiments which he had made with " Mus 

 rattus" and " Mus alexandrinus" ; by the latter name Rafinesque's 

 frugivorus and not Geoifroy's alexandrinus seems to have been im- 

 plied. Among the progeny of the various crosses effected, were 

 some peculiar rats which de I'lsle called " semi alexandrines^' ; judg- 

 ing from the description these must have corresponded rather 

 closely with Geoffrey's alexandrinus in outward appearance. De I'lsle 

 demonstrated that rattus, fntgivoms, and alexandrinus are nothing 

 but colour phases of one and the same species, viz., R. rattus, Lin- 

 naeus. His experiments suggested that the wild-coloured frugivo- 

 rus represents the primitive stock, properly belonging to warm 

 temperate or sub-tropical regions ; and that the dusky coloration of 

 typical rattus is simplj'- a change of hue brought about by the indoor 

 life' forced upon the species by its successful endeavours to colonize 

 colder lands. The species appears to have made its way to north- 

 western Europe at about the time of the Crusaders ; and by the 16th 

 century, at the latest, it had fullj'' assumed there its familiar dusky 

 garb. Geoffroy's alexandrinus may be regarded as an intermediate 

 stage, the belly having acquired within doors a darker colour, and 

 having lost its sharp contrast with the flank tint, although dorsal 

 darkening has not taken place to any considerable extent. In 

 examining a large, cosmopolitan collection of rats, it is quite 

 easy to find and arrange a series of individuals connecting 

 frugivorus with alexand/rinus, and especially the latter form with 

 typical rattus. 



Mendelians, as Bonhote (P.Z.8., 1910, p. 653 and 1912, p. 6), 

 argue that these three forms of rattus have arisen as mutations. 

 There is nothing inconsistent between this view and the history of 

 typical rattus as outlined above. In any case the colour differentia- 

 tion in the three races is susceptible of a physiological explanation. 

 Recently the three European races have been treated as 

 subspecies, the characters of which may be tabulated as follows : — 

 A^ — Dorsal parts black. 



Ventral parts dusky ; the hairs on belly 



short and usually adpressed ... ... R. rattus rattus, Linn. 



B. — Dorsal parts brown. 



a. Ventral parts not usually conspicuous- 

 ly lighter than flanks ; ventral hairs 

 with slaty bases ; coat harsh and usiial- 

 ly thin ... ... ... . . ii, rattus alexand/rinus, 



Geoff. 

 h. Ventral parts light coloured, sharply 

 contrasted with flanks ; ventral hairs 



