354 VICTOR E. SHELFORD. 



and gave us important elements of progress. Livingstone, 

 Brown, Transeau, and Shimek have made comparative studies 

 of the rate of evaporation in different habitats. Fuller most 

 recently has studied the rates of evaporation in well understood 

 stages of forest development which we discussed in the preceding 

 paper. The sum total of evidence at hand indicates that the 

 laws of succession and the physical conditions on the one hand 

 and growth-form and mores on the other are very generally in 

 accord. 



Ecological classification of animals must be based upon com- 

 munity of physiological makeup, behavior, and mode of life. 

 Those natural groups of animals which possess likenesses are the 

 communities which we must recognize. One community ends 

 and another begins where we find a general more or less striking 

 difference in the larger mores characters of the organisms con- 

 cerned. 



7. Ecological Terminology. 



Terminology in ecology is still unsettled and changing. Group- 

 ings have thus far been based upon similarity of habitat. Habitat 

 likenesses have, in general, been based upon general resemblances. 

 General resemblances have not always been accompanied by 

 similar physical conditions, as was pointed out in the preceding 

 paper of this series. In general, there has been an agreement in 

 the recognition of strata, of associations as communities based 

 upon the minor differences in habitat, and formations based upon 

 the larger major differences in habitats. Dahl ('08) uses the term 

 zootope for formation and bioconose for association and appar- 

 ently stratum also. Clements uses consocies for a division of a 

 community dominated by some one species of plant; the term in 

 this sense is less applicable to animals than to plants. 



We give the communities of different orders below with taxo- 

 nomic divisions of corresponding magnitude opposite, for com- 

 parison. With the exception of the first, these taxonomic 

 groupings do not bear the slightest relation to the ecological 

 groupings, but are added to indicate magnitude. 



Mores (the term applied to animals possessing certain attri- 

 butes) are groups of organisms in full agreement as to physio- 



