IN DETAIL 299 
nite phenomena of alternation which can be found anywhere. The family 
and the community, though the latter is less distinct in outline, are equally 
valid structures, the proof of which anyone can obtain by thorough methods 
of study. 
348. Nomenclature of the divisions. The suffix -efum is used to desig- 
nate a consocies of a formation, e. g., Picetum, Caricetum, etc. When two or 
more species characterize the area, the most important, or more rarely, the 
two are used. The termination used to designate a society is -ile, as Asterile, 
Sedile, Rosile. The suffix which denotes the community is -ere, and for the 
family, it is -on, viz., Giliare, Bromare,, Bidenton, Helianthon, etc. Layers 
are indicated by the affix -anim, as Opulasteranum, Verbesina-Rudbecki- 
anum, etc. It is evident that these suffixes, like the terms to which they re- 
fer, must be used always for the proper divisions if they are to have any 
value at all. There has been a marked tendency, for example, to use -etum 
in connection with the names of groups of very different rank. It is hardly 
necessary to point out that such a practice does not promote clearness. The 
following tabular statement will illustrate the application of both terms and 
suffixes : 
Picea-Pseudotsuga-hylium formation(-ium) Paronychia-Silene-chalicium 
Picetum consocies (-etwin) Paronychietum . 
Opulaster-Ribesanum layer (-anum) 
Opulasterile society (-ile) Androsacile 
Thalictrare community (-are) Festucare 
Pirolon family (-on) Arenarion 
349. The investigation of a particular formation. A comprehensive and 
thorough study of a formation should be based upon as many examples of it 
as are accessible. The example which is at once the most typical and the 
most accessible is made the base area. This plan saves time and energy, re- 
duces the number of instruments that are absolutely necessary, and estab- 
lishes a common basis for comparison. The inquiry should be made along 
four lines, all fundamental to a proper knowledge of the formation. These 
lines are: (1) the determination of the factors of the habitat, (2) a quadrat 
and a transect study of the structure of the formation, (3) a similar inves- 
tigation of development, (4) a floristic study of the contiguous formation, 
with special reference to migration. The sequence indicated has proven to 
be the most satisfactory, and is to be regarded as all but absolutely essential. 
Naturally, this applies only to the order in which the various lines are to be 
taken up, as they are carried on together when the work is fully under way. 
Since instrument and quadrat methods have already been given in detail, it 
