354 Harper : Native weeds and their probable origin 



ated, and that the multiplication of clearings has given these plants 

 opportunity to become widely disseminated. The chief objection 

 to this theory is that the only naturally treeless areas in New 

 England and New York (as far as can be ascertained from botan- 

 ical literature) seem to be bodies of water, marshes, meadows, 

 sand-plains, dunes, high cliffs, and mountain summits ; * and few 

 if any of the plants in question belong to such habitats. But 

 even granting that there could have been natural openings of a 

 kind which no longer exist in their original form, if all our native 

 weeds inhabited them their flora must have been far richer in pro- 

 portion to area than that of any other habitat known in the east- 

 ern United States today, which is incredible. 



It would seem therefore that the only possible explanation for 

 most of these species is that they have originated in some manner 

 or other in the last few centuries, as I ventured to suggest in 

 case of some Georgia weeds not long ago.f Some of the species 

 in question may even have originated within the last fifty years or 



the 



so, for it is scarcely conceivable that all the new species of Aster 

 Viola, Crataegus, Rubus, etc., which have been described from un- 

 natural habitats in recent years could have been overlooked by the 

 botanists of a century ago, if the plants existed then. Our herbaria 

 do not help us much in this particular, for specimens more than 

 fifty years old are very rare, and almost never cited in the descrip- 

 tions of new species. 



There is more than one way of accounting for the modern 

 origin of some of these species. Dr. Brainerd % has sagely sug- 

 gested that the great multiplication of species of Viola, Crataegus, 

 Panicum, etc., is chiefly due to increased opportunities for hybridi- 

 zation; which seems to be a modification of Kernels § theory that 



* According to Thoreau the only openings in the Maine woods, about the middle 

 of the 19th century, were rivers and lakes, burns and clearings, mountain summits, an 

 perhaps a few meadows. The few plants of this class mentioned by him grew in 

 burns and clearings. There are some evidences of about two square miles of dry Ian 

 naturally treeless among the Adirondacks in Hamilton Co., N. V., i,«55 feet ab ° VC 



. v „„^» «muug ujc r\uironaacKS in Hamilton v_,o., in. ».» ») w :j * 



sea-level (see Field & Stream 12 : 490. Oct. 1907, and the West Canada Lakes 

 pographic sheet of the U. S. Geological Survey), but there seems to be nothing 

 record about its flora. 



f Ann. N. Y. Acad. Sci. 17 : 116. 1906. 

 X Rhodora 8:10. 1906. 



I Pflanzenleben 2 : 582-588. 1891. (English translation, " Natural HW*J 

 Plants 7 ' 2 : 595-600.) 



to 



on 



of 



