660 TRANSACTIONS OF SECTION K. 
subjected, look unlike corresponding fresh material of the same or similar 
plants, so may trade samples, owing to special treatment, bear little outward 
resemblance to the same organs and tissues when fresh. 
When material of economic plants is ample another difficulty may be 
encountered. Domesticated species often undergo perplexing variation. In 
studying this variation the systematist may have to seek linguistic and 
archeological help, and be led into ethnological and historical by-paths. In 
classifying the forms that such domesticated plants assume he gladly avails 
himself of aid from those whose capacity for detecting affinities is unusually 
developed. But even with extraneous assistance the systematist, in this field, 
sometimes fails to attain final results. He can, however, always pave the 
way for the student of genetics, whose work involves the study of the ‘species’ 
as such. As regards forms of economic importance which neither organo- 
graphy nor anatomy can characterise, but which the chemist or biologist can 
discriminate, physiological methods are required to explain the genesis or 
elision of qualities evoked or expunged under particular conditions. 
A highly developed capacity for aggregation, if properly controlled, is also 
useful in the study of plant distribution from a physiographical standpoint. 
The systematist shows his sympathy with phytogeographical needs in two 
very practical ways. He declines, out of consideration for the geographical 
botanist, to deal with inadequate material, and for the same reason he refuses, 
in monographic studies, to be influenced by geographical evidence. The mono- 
grapher is conscious that if he pronounces two nearly related forms distinct, 
merely because they inhabit two different areas, he is digging a pit into which 
the phytogeographer may fall when the latter has to decide for or against a 
relationship between the floras of these two tracts. But the fact that, with 
existing knowledge, uniform delimitation of species is impossible, seriously 
weakens the value of normal systematic results for phytogeographical pur- 
poses. The units termed ‘species’ that are most useful in floristic and 
economic study are often too finely cut to serve distributional ends. When 
all existing plant-types have been treated on monographic lines the results 
may with relative safety be used by the phytogeographer, since errors due to 
personal equation may be regarded as self-eliminating. As matters now 
stand, however, the geographical botanist obtains his evidence partly from 
monographs, partly from floras, and is apt to be misled. Yet even in 
floristic work the systematist sees that the ‘species’ which it is his duty to 
recognise often arrange themselves in groups of nearly allied forms, These 
groups, which need not be entitled to sectional rank, while very variable as 
regards the number of species they contain, are more uniform than species in 
respect of their mutual relationships. They are therefore more useful than 
species as units for phytogeographical purposes. In defining these groups 
the faculty for aggregation is essential, and those in whom this faculty is 
highly developed may here be profitably employed, even when their discrimi- 
nating powers show a certain amount of atrophy. 
The cases, by no means rare, of workers who, with a comparatively poor 
eye for species, display great talent in their treatment of genera, afford 
indirect but striking proof that the faculty for aggregation may be more 
highly developed than its complement, and that the dominance of this faculty 
may ensure useful results. But the a priori expectation that in dealing with 
families this dominance should be still more valuable is not borne out by 
experience, for in this case it is recognised that aggregation has probably 
been pushed too far. This error has not been attributable to the faculty for 
aggregation so much as to the evidence at its disposal ; the corrective has 
largely been supplied by the use of anatomical methods as supplementary to 
organographic data. 
The physiologist in studying processes is not always obliged to take account 
of the identity of the plants which are their theatres of action. He has at 
hand many readily accessible and stereotyped subjects whose identity is a 
