RATTUS 733 



(a) Size large (head and body to 112, hind -foot to 25 mm.); dorsal 

 colour not rufous. 

 (a^) Skull slightly smaller (condylo-basal 

 length to 24.7 mm.), ridges feeble, 

 brain-case smoothly rounded anteriorly. 

 (fl2) Ventral surface dark, pectoral spot 

 often lengthened into a median 

 longitudinal thoracic and abdominal 

 streak. ...... hebridensis kebridensis 



(Lewis). 

 {b'^) Ventral surface silvery, pectoral spot 



absent or feebly developed. . . hebridensis maclean 



(Mull). 

 {b^) Skull larger (condylo-basal length to 

 25.5 mm.), ridges comparatively strong, 

 brain-case angular anteriorly. Belly 



silvery ; pectoral spot evident. . . hebridensis hamiltoni 



(Rum). 

 (^) Size smaller (head and body to 95, hind- 

 foot to 23 mm.). General dorsal colour 

 rufous. Belly silvery ; pectoral spot scarcely 

 developed. ...... hebridensis cumbrae 



(Great Cumbrae). 



Genus RATTUS { = Epimys). 



On p. 575, et seq., Epimys Trouessart is used as the generic 

 name of the true rats. Hollister {Proc. Biol. Soc, Washington, 

 19 1 6, 29, p. 124) has pointed out, however, that Fischer in 

 1803 {Das National Museum der N aturgeschichte zu Paris, 

 Bd. 2, p. 128) used Rattus (misprinted Ruttus) for this 

 purpose validly. The name Rattus must therefore supersede 

 Epimys, a regrettable though quite unavoidable change. 

 Hollister thought that decumanus (i.e., norvegicus, was the type 

 species of Fischer's genus Rattus ; but Thomas {Ann. Mag. 

 Nat. Hist., 18, p. 240), and Hinton {Journ. Bombay Nat. 

 Hist. Soc, 23, p. 59) agree that Fischer took rattus and not 

 norvegicus as the genotype. The correct technical names 

 of the species occurring in Britain are now therefore : — 



1. Rattus rattus, with its sub-species r. rattus, r. alex- 

 andrinus, and r. frugivorus. 



2. Rattus norvegicus. 



