356 VICTOR E. SHELFORD. 



wood of living trees, probably, should be considered as consocies 

 relatively independent of stratification phenomena. 



Associations are groups of strata uniform over a considerable 

 area. The majority of mores, consocies, and strata, are different 

 in different associations. A minority of strata may be similar 

 though rarely identical. The unity of associations is dependent 

 upon the migration of the same individual and the same mores 

 from one stratum to another at different times of day or at 

 different periods of their life histories. Such migration is far less 

 frequent than from one association to another. 



Formations are groups of associations. Formations differ from 

 one another in all the strata, no two being closely similar. The 

 number of species common to two formations is usually small 

 ( e - -i 5 P er cent.). Migrations of individuals from one formation 

 to another are relatively rare. 



To illustrate associations and formations, we have noted two 

 great groups in the forest development series discussed. These 

 groups are the cottonwood, pine, blackoak associations belonging 

 to the sand area and in disagreement in the majority of mores, 

 consocies, and strata; and the redoak, hickory, and beech asso- 

 ciations belonging to the climatic forest proper and comparable 

 with the first group in disagreement. The mores of the former 

 are characteristic of sand areas, within the range of the deciduous 

 forest climate. On this basis we may designate this as the sand 

 area animal formation of the deciduous forest climate (Gleason, '10). 

 It is here made up of the three associations just mentioned. The 

 three later stages constitute the deciduous forest animal formation, 

 which is here also made up of the three animal associations, named 

 above. The two formations are separated upon the basis of 

 striking differences in modes of life of the animals of the sand- 

 dominated and forest-dominated communities and are in general 

 disagreement as to mores, consocies, and strata concerned, only 

 about 5 per cent, of the animals of the two hundred species listed 

 in the preceding paper occurring sparingly or occasionally in the 

 more similar associations of the two formations. The character 

 of these differences was briefly outlined on pages 89-90 of the 

 same paper (Shelf ord, 'I2 1 ). 



Extensive or climatic formations are groups of formations 



