323 



extremes of variatiou met with at single localities are as great as those 

 from the most widely separated of the above-named localities; in other 

 words, no geographical variation in size is perceptible. The largest 

 northern specimen (canadensis), from Peel River, Arctic America, with 

 a length of 5.30, a little exceeds in size the largest specimens from any 

 locality south of the latitude of 40° ; but it in turn is slightly smaller 

 than a specimen (fasciatus) from Fort Townsend, Wash., which has a 

 length of 5.50, and by anotlier of the same dimensions [riifus) from the 

 Big Sioux Eiver. Eight specimens of the most southern type (L. maeu- 

 latus), all from Texas and the Mexican side of the Lower Eio Grande, 

 differ in the average from nine specimens of the most northern type [L. 

 canadensis), all from Arctic or sub-Arctic America, almost inapprecia- 

 bly, the canadensis series having an average length of 5.01 and the macu- 

 latus series of 5.00 ! The difference in breadth is also only about one- 

 tenth of an inch, which the addition of a single specimen to either series 

 might cancel. This is certainly a surprising result when it is remem- 

 bered that one of the chief alleged distinctive characters of L. cana- 

 densis has been its supposed larger size ! 



The average dimensions of these several series are as follows : 



Species. 



Number of 

 specimens. 



Length. 



Breadth. 





9 



8 

 10 



5.01 

 5.03 

 5.00 

 4.91 



3.52 





3.56 





3.40 



*L. nifus .. . . 



3.41 







Mean of all 



34 



4.98 



3.47 



ft 





* The specimens placed nnder rufus are those that are so marked in the coUection, being the speci 

 mens so identitied by Professor BairJ. 



The fasciatus series is the largest, but this series happens to include 

 more very old specimens than the others, and hence its higher average. 

 Such a constancy of size as is here shown to prevail over an area 

 embracing more than 40 degrees of latitude is probably without a par- 

 allel in any other conspecihc group of Korth American Mammals. 



The difference bet ween these heretofore com monly recognized " species'' 

 of the genus Lynx must hence be sought elsewhere than in size. The 

 specific distinctness of L. canadensis, the most northern type, has been 

 heretofore scarcely questioned, in consequence of its supposed larger 

 size, larger limbs, longer, softer pelage, longer ear-tufts, more indis- 

 tinct markings, and generally lighter or grayer color. The. longer ear- 

 tufts correlate with the longer, softer pelage, that always characterizes 

 the boreal representatives of species having a wide latitudinal range. 

 The difference in coloration is not greater than, or even so great as, that 

 which obtains between fasciatus and rufus, or between fasciatus and niacu- 

 latus, which forms natnralists now seem disposed to refer to one 

 and the same species under the name L. rufus. Maculatus, the 

 most southern form, differs from the "typical" or eastern rufus in its 

 shorter, coarser fur, more reddish tints, and more distinct markings. 

 Its reputed range extends from the Lower Kio Grande westward across 

 the continent to Southern California ; but in the National Museum col- 

 lection are also specimens marked rufus from many points within this 

 area, including a considerable series from Fort Tejon. The gradation 

 from the " typical" rufus type into maculatus is complete and by almost 

 insensible stages. 



