200 MONOGRAPHS OF NORTH AMERICAN RODENTIA. 



in this memoir. Some specimens are little more than half as long as others, 

 and certainly have less than half the hulk. Our figures for apparently 

 mature animals run from 4.50 to 8 inches Cor length of trunk. Now, to 

 keep largely within hounds, we will strike off half an inch each way, and 

 say a normal limit of variation is between 5 and 7£ inches: this 2J inches 

 is 50 per cent, of the minimum and 40 per cent, of the mean length of the 

 animal. This great discrepancy is the more instructive, because in the case 

 of xanthognathus there is no possible; question of specific identity of the 

 largest and smallest specimens. In some other eases, where reputed nominal 

 species, based in part on dimensions, were at issue, we may possibly be 

 suspected of granting improbable and undue range of variation. But here 

 the matter is brought to a focus: we show, in the specimens of unquestion- 

 ably a single species, as great variability in size as we have anywhere 

 attempted to prove. 



And yet this difference is no greater than we believe is well known to 

 occur in other species of the genus, notably the Arvicola amphibius of 

 Europe. No one is surprised to kill two house-rats, one of which is twice 

 as big as the other. We hold that a corresponding variability is as normal 

 to some purely feral animals as to the semi-domesticated species just cited; 

 and we believe that it argues a progressive increase in size, with age, over 

 the stature ordinarily reached at the period of puberty — that is to say, a 

 Mus or an Arvicola may be "adult" or "mature" in the sense that it has lost 

 the signs of youth, gained those of adult life, and become capable of repro- 

 duction, and yet, after this, may increase in length by one-third at least, and 

 double its bulk in the subsequent years of its life. 



Recurring again to our measurements, we next observe that the tail of 

 this animal (taking it to the end of the vertebrse as a more constant and 

 reliable measurement than to the tip of the hairs) ranges from 0.75 to 2.25, 

 as the figures stand! and, making large allowance for erroneous elements, we 

 may safely say that the tail is " an inch or tiro" long, i. e., it varies 100 per 

 cent, of the minimum ! What could more forcibly illustrate the instability 

 that attends the dimensions of organs produced in any sense as matters of 

 vegetative repetition ! The measurements of other parts need not detain us. 

 For several reasons, among them ease of correct measurement, the limits of 

 the figures for the feet and ears do not stand quite so far apart as (hose for 

 the body and tail do; they coincide with the results of our measurements 



