ECOLOGY OF ISLE ROYALE. 165 



"Thus, in both Coleoptera and Orthoptera, the difference between the two zones 

 is seen to be very marked, not only as to species, but also as to genera, show- 

 ing that we have to deal with distinct fauna, p. 312. 

 ^ "So far as I am able to judge, the suppression of the central region is en- 

 tirely justified, but I cannot agree as to the proposed Sonoran region. An 

 analysis of the insects of the Colorado Rocky Mountains shows that the high- 

 alpine and mid-alpine elements, although sufficiently distinct, are both essential- 

 ly boreal. If we follow Dr. Merriam's arrangement, it appears that the high- 

 alpine is truly boreal while the mid-alpine belongs to the transition region, 

 containing a considerable number of strictly American types. The sub-alpine, 

 on the other hand, is southern or Sonoran. 



"Dr. Horn has kindly given me his opinion as follovra: 'My ideas of the 

 distribution of the Coleoptera in the mountainous region of Colorado, which 

 is a good center of the Rocky Mountain chain are as follows: The high region 

 seems to have been populated from the Canadian through the H. B. T. region. 

 A collection made above 8,000 feet in Colorado is almost identical with one 

 made in the Lake Superior region. The same fauna runs down to N. M. and 

 Arizona, and again appears, mixed, of course, in the Mexican Mountains. 



" 'The sub-alpine region is one that continues from Washington to New Mexico, 

 as shown by such striking forms as Ergates, Melanophila miranda, Iphthimus 

 serratus, Oaleruca externa, Calosoma lunatum, in varieties. 



" 'The lower region, foot hills, etc., is a mixture of New Mexico forms with 

 those of the Eastern United States, with some peculiar forms allied more to 

 the southern regions. 



"'California is a peculiar region, and, in many respects, allied to Europe (in 

 general). I think California supplies us with more species of genera peculiar 

 to Europe than does the Eastern region.' (In litt., July 14, 1892.) 



"According to the facts now recorded it seems that there is, firstly, a circum- 

 polar and strictly boreal element; secondly, a boreal but modified or Canadian 

 element; and thirdly, a southern element belonging to the arid portion of Dr. 

 Merriam's Sonoran region. I do not thinK any distinct faunae except these 

 can be recognized, and the central region acciordlingly falls. But there is, 

 sprinkled among the ordinary types, a distinct element of endemic species, to 

 which I shall refer later. There also seems to be a few surviving fragments 

 of an ancient fauna, of which Anthracopteryx is a good example. 



"There seems to be a small California element, but the species falling under 

 this head are perhaps rather Southern than properly Californian. pp. 313-314. 



"The resemblance between the Colorado fauna, and that of the Mississippi 

 basin and further East, always, excepting the boreal element that comes from 

 the North, is very slight indeed. The great plains to the east of the Rocky 

 Mountains have been as much a barrier as the sea would have been. p. 314. 



"A Method for Defining Faunal Regions. It appears from a consideration 

 of what has been written on faunal regions, that it would be desirable if some 

 rules could be laid down, so as to leave the matter less to the discretion of 

 the individual writer. It would require a good deal of research to determine 

 what rules could be laid down, that would work, but as regards insects, at 

 all events, I have thought it possible that the following rule might answer for 

 secondary faunal divisions: 



"Any tvx) districts shall be regarded as in the same secondary faunal divisions 

 if the number of species common to both exceeds the number of genera in com- 

 mon, p. 315. 



"Equigeneric Areas. For minor divisions, to be used in relation to particular 

 groups, I have devised what may be termed equigeneric areas. 



"Equigeneric areas are areas throughout which the genera of the group under 

 consideration are identical. 



"These areas are sometimes large, sometimes small. When two genera over- 

 lap, the region where they both occur, however small, makes a separate equ- 

 generic a.rea. This might be thought a disadvantage; but really, I believe it 

 to be an advantage in the method, since it is important to recognize these 

 intermediate or overlapping areas, p. 316'. 



"Origin of the Rocky Mountain Fauna. The numerous fossils of Colorado 



bear testimony to the fact that the region of the Rocky Mountains has in the 



past been peopled by a highly remarkable and numerous fauna. This fauna, 



however, does not appear to be ancestral to that of the present day. Nor has 



22 



