HISTORICAL 105 



wahrend andere parallel neben einander laufen. Auch 

 schon vor Ausbildung einer Schlussformation trifft nicht 

 selten eine Reihe mit einer anderen zusammen und geht in 

 deren Richtung weiter. ' ' 



Pound & Clements ( J 898: 216, 259) traced the development 

 of the blowout formation and of waste formations on denuded 

 areas. They distinguished formations as primitive or 

 recent with respect to origin ; the latter alone are to be con- 

 sidered in the study of existing successions. Recent 

 formations arise by nascence, as in the case of areas desti- 

 tute of a floral covering, or by modification, when a patch or 

 formation is changed by the invasion of other species. 

 Cowles' work (1899:95) on the dune vegetation of Lake 

 Michigan, modelled after Warming's studies on the Den- 

 mark coast, was the first special account of succession to be 

 published in this country. The author has classified dunes 

 as embryonic, wandering and established, and has followed 

 the detailed development of vegetation in each of these. 

 His conclusions form an excellent r^sum^ of the stages in 

 dune successions. Nilsson (t899:89) followed in a detailed 

 manner the development of vegetation on the cliffs and in 

 the swamps of middle Sweden. The cliff succession is 

 especially interesting on account of the full treatment of the 

 hchen stages. Cowles (1901:73) has applied his methods of 

 successional research to the study of river-valleys and shore- 

 lines with very instructive results, and he has summed up 

 his investigations in a developmental classification of much 

 value. Whitford ( J90h289) has made a study of the forests of 

 northern Michigan along the same lines. Weber (J90J , 1902) 

 has made further contributions to the developmental history 

 of moor and forest, Graebner (1902) to the phylogeny of the 

 heath vegetation and Clements (1902) to the succession of 

 talus formations. 



Succession is the phenomenon in which a. series of in- 

 vasions occurs in the same spot. It is important, however, 

 to distinguish clearly between succession and invasion, for, 

 while the one is the direct result of the other, not all in- 

 vasion produces succession. The number of invaders must 



