1909.] R. E. Lloyd : The Races of Indian Rats. 5 



agree in showing some slight peculiarity distinguishing them from 

 other species. Whether it is possible to rediscover at a later date 

 species defined in such a manner seems doubtful ; and if it seemed 

 possible, the question as to whether the individuals on which the 

 rediscovery was alleged would be genetically related to those on 

 which the species was originally defined, appears even more 

 doubtful. 



In the writer's opinion it is not possible to "classify" the 

 Oriental rats in a satisfactory manner on the lines on which classi- 

 fication is being attempted at the present day. Before commenc- 

 ing systematic work in a group, the observer must make himself 

 acquainted with the structural features which have been chosen 

 by previous workers in defining new species. In the case of the Mus 

 rattus group the following points are considered of sufficient weight 

 to distinguish species: a difference of 30 mm. in " length " ; a 

 difference of 5 mm. in the length of the hind foot ; of 3 mm, in ear- 

 length ; a difference of 20 in tail percentage; as regards colour, 

 redness, yellowness or blackness are considered of importance ; 

 a white ventral surface is of some importance; a bicoloured tail 

 (dark above, white below) is regarded as of the utmost importance. 

 In 1903 Oriental rats were arranged in 96 species. 



If, with these points well in mind, the observer examines 500 

 rats caught in different parts of an eastern town and attempts to 

 classif}' them, he is confronted with an impossibility. (I do not 

 refer to mole-rats, bandicoots and Mus decumanus, which are 

 distinct at a glance.) He finds that variation among the hetero- 

 geneous collection is wide enough to embrace all, or nearly 

 all, the points chosen by previous writers on which to define 

 species ; but he also finds that ten rats caught in the same house 

 will often resemble one another verv closely in colour, size, and pro- 

 portion, always more closely than ten rats taken haphazard from 

 different parts of the town. The rats of two separate houses will 

 sometimes show differences as great as those chosen to distinguish 

 species, while the rats in each house resemble one another very 

 closely indeed. The following difficulty therefore presents itself. 

 Should any or all of these small groups receive specific names ? 



The common procedure which leads to the definition of a new 

 species of Oriental rat seems to be as follows : — 



A field naturalist sets traps in a rat-frequented house or close 

 to a set of burrows in a field ; two or three house or field rats are 

 caught ; in either case the specimens which closely resemble one 

 another come into the hands of a specialist in Europe, who 

 examines them regarding such features as those enumerated above. 

 Although they resemble one another very closely, it is highly 

 improbable that they will, in every way, closely resemble any 

 previously described species. They are defined as a new species. 

 The discover}'^ of new species on these lines is slow because the 

 number of rats examined is quite insignificant when compared 

 with the number of rats present in one Oriental town. Though 

 slow, the process cannot end. 



