igog.] R. E. Lloyd : The Races of Indian Rats. 95 



habitation to either Mus fulvescens or Mus niveiventer. Such forms 

 as Goalunda, Vandaleuria and Mus mettada are considered to be 

 quite distinct from the Mus rattus type. 



If we consider in succession the numerous single sports, the 

 small family groups of such sports (exampled b}' cases 5, 6 and 7), 

 the larger groups (case 8), and the established races i — 4, they 

 seem together to illustrate the process by which a new race arises 

 from an old one. 



During one week at least four well-marked sports were met with 

 in Rangoon alone, whereas the established races of rats in India 

 are very limited in number. Few sports can establish a family 

 group, ver}' few family groups can establish a race. One cannot 

 do more than guess at the nature of the ' ' fitness ' ' which renders 

 one sport successful and eliminates a dozen others. However, one 

 can den}^ that all successful sports are so b)^ virtue of the characteris- 

 tic features which are obvious to our eyes. A short bicoloured 

 tail, a long bicoloured tail, a sixth palatal ridge, a tufted tail are 

 merely the marks of successful races, the}'' are surely not the features 

 which determine the success. The factors which determine the 

 success of a group of rats may be a greater pugnacit}' or cunning, a 

 greater resistance to certain diseases ; and a greater fertility, com- 

 bined with that mysterious factor prepotence ; or they may be in- 

 conceivable. 



Hitherto we have been considering races which show com- 

 paratively' small differences from one another. There are in the 

 fields round Amritsar two races of mole-rats living side by side 

 always in separate colonies or sets of burrows. These show such 

 structural differences from one another that the}' have received 

 different generic names ; one of these, Gunomys, is a genus which 

 occurs throughout the whole of India. The other, Nesokia^ seems 

 to be confined to the north-west. Gunomys has a tail percentage 

 of 80, small incisor teeth, a long palatine foramen and a large num- 

 ber of mammae ; it produces 8 — 12 young at one birth. Nesokia 

 has a tail percentage of 50, large incisor teeth, and concomitanth* 

 a very short palatine foramen ; few mammae, and it produces 2 — 4 

 young at a birth. The skulls of these two genera are shown 

 on plate iii. It may be safely assumed that Nesokia was derived 

 from Gunomys. 



If we suppose that it was derived from Gunomys b}' ' ' natural 

 selection " working on normal variation, we must assume that there 

 were a number of steps linking together the two extremes. The 

 steps — a series of hypothetical species now extinct — must have each 

 shown, gradually and in turn, a reduction in the length of the tail 

 and palatine foramen, fewer mammae and young. Bach of these 

 hypothetical species must have fitted its environment and must 

 have become changed in response to some slight alteration in that 

 environment ; although at the present day the two extremes are 

 perfectly suited to what appears in our eyes to be the same environ- 

 ment. 



