1907.] Dr. Hossack: The Rats of Calcutta. 23 



instances have been excluded in making up these figures, first a A^. bengalensis skull 

 with nasals 10 per cent, below the average, and a smallest size M. rattus skull 6 per 

 cent, above the average. The coronoid suture shows a considerable amount of varia- 

 tion A simple flattened curve is common, but it is still more common to find that 

 the curve is like a bracket owing to a posteriorly projecting point on the frontal. 

 In 6 out of 62 skulls examined on this point it was angular, the two Hmbs forming an 

 angle of 120°. This condition is almost always found in immature skulls ; it is not 

 quite sufficiently marked in the young skull as figured in fig. 5. In four specimens 

 the posterior projection of the frontal was very marked. Number 259 had a zygo- 

 matic plate that sloped slightly forward and downward instead of being perpendicular. 

 This sloping, non-developed condition of the plate is marked in very young skulls 

 with the third molar not yet developed. Number 182 showed the right anterior palatine 

 foramen 75 mm. longer than the left. Close to the coronoid suture there is a projec- 

 tion of the supraorbital ridge, from which a vertical ridge runs down the posterior 

 wall of the orbit ; in No. 284 this is almost completely absent. The interparietal, on 

 which much stress is laid as a means of differentiation, varies very much in shape. 

 It is commonly an oval with pointed ends. The anterior curve is generally bracket- 

 shaped owing to an anterior projection ; the posterior curve may be similarly modi- 

 fied by a posterior projection. I^astly, the ends of the oval may be truncated so that 

 the bone has lateral boundaries and tends to be five-sided. As a rule the posterior 

 suture is clear of the occipital ridge, but sometimes it is not, so that the interparietal 

 has a considerable share in the formation of the ridge, vide fig. 31. 



The occipital in Nos. 91 and 132 instead of being vertical, slopes slightly backwards 

 and downwards, though not to the same extent as in Mus decumanus. Occasionally 

 the occipital ridge projects posteriorly beyond the vertical plane. 



Dimensions. — -It is not necessary to say anything as to the measurements of the 

 body and the cranium, as the tables given at the end have been broken up into groups 

 which indicate clearly what are the average normals and what are above or below these. 

 The two smallest sizes have not been taken into account in making up the averages 

 of body and skull measurements given in the tables of comparative measurements at 

 the end of this paper. 



MUS RATTUS var. NITIDUS. 



Thomas finds the skull in no way different from that of M. alexandrinus or M. 

 rufescens and makes it an intergrading variety — the hill variety of M. alexandrinus. 

 From an examination of a very limited number of specimens in Darjeeling, it seems to 

 me to be a distinct variety clearly marked off by the proportionate length of the tail. 

 In two adults this was 97 3 of the length of the head and body, and in two three- 

 quarter grown rats 1057 P^^" cent. The tail is distinctly lighter below, unlike that 

 of M. rattus as found in Calcutta. Blanford makes the tail of M. rattus generally the 

 same colour throughout, but sometimes paler beneath, and states that some specimens 

 from Simla have the undersurface of the tail quite white. Possibly the real explanation 



