1397.] Botany. 983 
It has been pointed out in a previous paper that the formation is a 
definite phytogeographical concept, determined by its statistic, vegeta- 
tion forms and habitat groups. The determination, or better, the 
recognition of a formation is conditioned by the previous analysis of 
its constituents from these various standpoints. This fact was first 
appreciated by the Continental phytogeographers, and Drude was the 
first to carefully elaborate the formations of an entire flora in accord- 
ance with this principle. The fundamental nature of his work in this 
line renders his researches classic, and his conclusions superlatively 
final, for the present at least, As is well shown in his Deutschlands 
Pflanzengeographie, I, it is necessary to first comprehend the floral 
covering by a careful analysis of its floristic, vegetation forms and hab- 
itat groups, and then by even more careful synthesis, determine the 
formations thus pointed out. While the floral covering may be truth- 
fully likened to a mosaic in which the various pieces are formations, it 
is not true that every distinguishable bit of difference is a formation. 
It must likewise be borne in mind that formations are by no means of 
fixed and absolute expression: they are at all times more or less plas- 
tic, manifesting modification, incipience, decadence, and most intimate 
relations with other formations. 
It is, at present, so common to reproach European botanists, and par- 
ticularly those upon the Continent, for their disregard of the results 
of American investigation, that it is odd to be obliged to call attention 
to American neglect of European research. Yet the article in hand 
manifests justs such a lack of acquaintance with the fundamental con- 
tributions of Drude. Drude’s concept of the plant formation is un- 
doubtedly to be regarded as the correct one. Measured by this stand- 
ard, the great number of formations noted for a small area by Profes- 
sor MacMillan fall in most cases to the rank of facies, or patches, or 
they are at best only incipient or decadent stages of real formations. 
The author has committed two serious mistakes in his elaboration of 
the Lake of the Woods formations: the first is lack of perspective, the 
second is too minute analysis. It is impossible to delimit formations 
accurately by studying the floral covering of a restricted area. Such 
limitation can be done intelligently and well only in a natural vegeta- 
tion area; except in the rarest cases, in nothing smaller than a vegeta- 
tion region. A formation must be studied in all its various stations 
and in its many phases, before final conclusions can be reached con- 
cerning its validity. From the lists given, the floral covering in the 
vicinity of the Lake of the Woods is composed of the most heterogenous 
elements, which have crowded together in peculiar fashion. Apprecia- 
