46 T O R R E Y A 



a fashion as a small-flowered variety of the Eurasian Alisuia Plantago-aquatica. 

 Not only are species and their subdivisions the product of opinions of indi- 

 viduals, but the same is true of the limits of genera and of higher groups. Xot 

 much is to be gained by a painful recital of the infinite variation of nomencla- 

 ture under present conditions ; it is much more illuminating to review the geo- 

 graphic conditions which have made or should make a background for nomen- 

 clature. 



This problem of glaciation in eastern North America has been ably treated 

 b_\' Professor Fernald. He has shown that many species of restricted distribu- 

 tion in Western Newfoundland, in the Gaspe Peninsula of eastern Quebec, and 

 in some areas adjacent to the Great Lakes, are ancient plants (in contradistinc- 

 tion to "\\'illis' "Age and Area"' hypothesis) that have persisted in places not 

 covered b}' the \Msconsin stage of the Pleistocene glaciation. The vegetation of 

 the glaciated area we may presume to have been obliterated during the ice age, 

 and since the deposits of the coastal plain are of comparatively recent origin 

 (chiefly marine) the uplands of the southern Appalachians and the Ozark 

 Mountains remain as areas from which the flora now inhabiting the coastal 

 plain has probably been derived. These various areas are shown in detail in 

 Fernald's'* recent work on the \'irginia coastal plain. 



Species which cover the three main areas (Appalachian uplands, glaciated 

 area, and coastal plain) often show marked divergences in structure in these 

 individual areas, and constitute geographic varieties, which if the variations 

 increased (according to the Darwinian interpretation mentioned in my open- 

 ing paragraphs ) might become distinct species. The Appalachian plateaus still 

 harbor many species of the coastal plain, and from my own observations on the 

 vegetation of the barrens of ^Middle Tennessee, it seems probable that such 

 plants as Panicuui meridionale, Rynclwspora macrostachya, Scleria reticularis, 

 and Eleocharis microcarpa have moved into the Great Lakes area from the 

 siliceous uplands of Tennessee and Kentucky through Indiana, as we may infer 

 from isolated occurrences in the last named state. The bicentric range of 

 Lilaeopsis carolincusis in south-eastern United States and in the Argentina 

 region of South America is also shown. A similar disrupted distribution is 

 common in other groups, especially in the Cyperaceae, and is well shown in a 

 number of species of Eleocharis. Whatever may be the geographical explana- 

 tion, the problem of correlating published varieties and other subspecific units 

 in variable species with such bicentric ranges is well-nigh insuperable; it is 

 perhaps the most cogent argument for the non-recognition of varieties. 



We may now turn attention to a recent publication of extraordinary interest 

 by J. C. Willis,^ the author of "Age and Area." This work, entitled: "The 



*Rhodora42: 367. 1940. 



= Cambridge University Press, England. 1940. 200 pp. Quotations bj- permission of Tiie 

 Macmiillan Compam% publishers, U. S. 



