OF SELBORNE. 45 



and will endeavour to put the matter out of doubt, 

 whether it be a nondescript species or not. 



I suspect much there may be two species of water- 

 rats. Ray says, and Linnaeus after him, that the 

 water-rat is web'-footed behind. Now I have dis- 

 covered a rat on the banks of our little stream that is 

 not web-footed, and yet is an excellent swimmer and 

 diver: it answers exactly to the Mus amphibius of 

 Linnaeus (See Syst. Nat.), which he says, " natat in 

 fossis et urinatur.' 1 I should be glad to procure one 

 " plantis palmatis." Linnaeus seems to be in a puzzle 

 about his Mus amphibius, and to doubt whether it 

 differs from his Mus terrestris ; which, if it be, as he 

 allows, the " Mus agrestis capite grandi, brachyuros" of 

 Ray, is widely different from the water-rat, both in 

 size, make, and manner of life 10 . 



10 Willughby was the originator of the confusion alluded to. He 

 described the water-rat as having its toes connected together by inter- 

 vening webs ; and his description was published by Ray in the Synopsis 

 Quadrupedum. Linnaeus, believing that such authorities were to be 

 relied on, admitted a rat-like animal, having its hinder feet webbed, into 

 the several editions of his Fauna Suecica ; placing it, in the first of them, 

 where its technical characters directed him, in the genus Castor. Sub- 

 sequently he associated it with the rats ; and referred to it as of doubtful 

 existence, as being perhaps inaccurately described, and as probably to 

 be referred to his Mus terrestris. There can now be no doubt that he was 

 correct in regarding the large rat with a hairy tail of moderate length, 

 which frequents ditches in the summer time, and swims and dives well, 

 and which has on these accounts acquired the name of amphibius, as 

 identical with the one described by him as the terrestris, as having the 

 same outward form and colours, and as being found in burrows : the 

 winter nest of the species is described by White in Letter XXVI. 

 Willughby's error must have been occasioned by his having assumed 

 from a certain habit that a certain structure which he regarded as indi- 

 cated by it must necessarily be coexistent with it: but he should not 

 have forgotten, even for an instant, that natural history is a science of 

 observation, and not of theoretical deductions. 



The Mus agrestis capite grandi, brachyuros, of Ray, is indeed widely 

 different from the water rat : it is the short-tailed field mouse or vole, 

 Arvicola agrestis, FLEM. ; the water rat, or rather water vole, being the 

 Arv . amphibia, DESM. The genera Arvicola and Mus do not belong even 

 to the same primary section of the rodents. E. T. B. 



