33 ° Pruiiu — A Review of the genera 
a variety (var. rudis ) of M. horridtda. This plant, as was pointed out 
eleven years ago, is the M. racemosa of Franchet (PL Delavay., 41) but not 
the M. racemosa of Maximowicz. Now that more complete material is 
available it seems advisable to treat this Yunnan and Szechuen plant as 
a distinct species of the group Acideatae. The writer has already pointed 
out that M. horridtda , with its variety racemosa , is virtually a Tibetan 
species which only casually overflows through some of the eastern 
Himalayan passes into the higher valleys of Sikkim and Chumbi. In the 
north-western Himalaya, from Kamaon to Kashmir, we find M. horridula 
replaced by a representative species, M. aculeata ; in the eastern Himalaya 
M. horridtda is, at slightly lower elevations, replaced by another repre- 
sentative species, M. sinuata. It now appears that in the corresponding 
area and at corresponding elevations in Szechuen and Yunnan M. horridtda 
is similarly replaced by this third representative species, which is described 
below as M. rudis. 
Taxonomy. 
In arranging the species of Meconopsis the most satisfactory method is 
found to be that of grouping those spegies which agree in the greatest 
number of characters. This is not difficult, because, while we find in this 
as in other high alpine genera that the distribution of individual species is 
usually limited, we also find a complementary tendency to the occurrence, 
in distinct areas where the general conditions are similar, of representative 
species. The groups that it is necessary to recognize are somewhat unequal 
in number of component forms ; these groups, rather than the actual 
species, probably represent the natural units of the genus. For further 
aggregation it is necessary to look for characters common to the various 
groups. Within the limits of the genus there is a wealth of striking differ- 
ential morphological features and physiological peculiarities. Thus some 
species have stems, now simple now branched ; others have none. Some 
have, others want a style ; some have, others are without a disk. Some 
have papaveroid, others have sanguinarioid corollas ; some have radiating 
divaricate, others have decurrent contiguous stigmatic lobes. Then some 
species are monocarpic, others are polycarpic ; some have crowns that 
persist, others have crowns that die down during winter. Some flower in 
the same season as the seed, others flower in one year, or in two years, or 
occasionally after more than two years from seed. The temptation to 
employ as primary one or other of these salient differential characters is 
considerable. But further examination of all of them in the light of the 
actual characters of the natural groups shows us that the subdivision they 
effect is never complete. They traverse the evidence afforded by the actual 
groups and show us that in taxonomy there is no place for either morpho- 
logical or physiological prepossessions. There is, however, one character, 
