77 



texture or composition and available water, may be such that a 

 series of successional changes may be halted for an indefinite 

 period at a point far short of the regional climax. Thus the 

 reviewer has pointed out, in his discussion of the plant societies 

 of Connecticut,* that along the crests of the trap ridges edaphic 

 conditions may be such that the ultimate forest is dominated by 

 oaks and hickories; similarly many pitch-pine forests in the Con- 

 necticut sand-plains may very likely represent the most meso- 

 phytic type of vegetation attainable under the existing soil con- 

 ditions. It seems important therefore that a careful distinction 

 be made between an edaphic climax which may be determined 

 by local conditions, and a regional climax, which is favored by 

 climate but can be attained only under favorable edaphic con- 

 ditions. While, therefore, in a sense the vegetation of the pine- 

 barrens is to be regarded as a primitive or pioneer type, it prob- 

 ,ably represents a remarkable example of a widespread edaphic 

 climax, widespread because the peculiar soil conditions with which 

 it is so intimately associated are likewise widespread. 



Several pages of the introduction are devoted to a consideration 

 of "the probable effects of the glacier on the coastal plain ex- 

 cluding the pine-barrens" and to the northward distribution of 

 coastal plain plants into Staten Island and Long Island. In the 

 latter connection it is assumed that an "avenue of migration" 

 must have existed in post-glacial times between these areas and 

 New Jersey, yet the necessity of assuming the existence of some 

 such former connection between eastern Long Island and the 

 adjacent mainland to account for the segregation in southern 

 New England of a very considerable group of coastal plain 

 species "seems doubtful. "f 



Under the head of climatic factors an attempt is made to cor- 

 relate the distribution of the flora within the area treated with 

 the length of the growing season. It is pointed out that the 

 number of days intervening between the last killing frost of 

 spring and the first one of autumn varies from 1 17-123 days in 

 the Catskills and the mountains of Pennsylvania to 220 days at 

 Cape May, a difference of more than three months. On an 



* Torreya 14: 177. 1914. 



t In this connection, see the reviewer's observations in Torreya 13 : 94-99. 1913- 



