1883.] Orchidee in the Natural System. 1251 
out the aid of insects, and the contrivances serving the same end 
are many and various. Those with Papilionaceous flowers we 
may consider as specially modified, and these include nearly 
three-fourths of the genera and a still larger proportion of the 
species. Laying special stress upon the seeds and means for 
cross-fertilization here, as we have before, we see good reason for 
calling the Leguminose highly organized plants. The legumin- 
ous fruit and the papilionaceous flowers, are two things found in 
no other family, and these two, or either one, will tell a legume as 
easily as the compact head of flowers will point out one of the 
Composit. 
While the Leguminosz are entitled to the highest rank in the 
Polypetala, the Composit hold the highest place of all, for the 
former fall below the latter because of the separation of the 
petals, the almost universal prevalence of a calyx, so profoundly 
modified in the Composit, the increased number of stamens, 
the smaller number of seeds produced, and their production in a 
pod instead of separately. 
Closely allied to the Leguminosz stand the Rosacez, but not 
so highly specialized, having simpler flowers, more numerous 
Stamens and other characters. Allied to both of these is the 
order Saxifragez, and to it the Ranunculacee, and to the latter 
the Umbelliferæ. In numerical strength and in distribution too, 
do we find these several families arranged in much the same 
order. Leguminosz with 6500 species, Rosacez with 1000, Um- 
bellifera with 1300, and Ranunculacee with 540. In distribu- 
tion the Leguminosz have the widest range and are especially 
abundant in the tropics, the Rosacez come next, the Umbelliferz 
third, being rare in the lowlands of the tropics, and the Ranun- 
Culacez are last, largely confined to temperate and arctic regions. 
Referring now to the Monocotyledons, we find that of all the 
orders the Orchidez is the largest, there being known at present 
tween 4500 and 5000 species. The order thus stands third 
from Composite. We find the species of orchids to have a very 
local distribution, but the family itself is very widely scattered. 
Their metropolis, as with the Composite, is in the tropics, espe- 
cially in America, and we find a large proportion of them to be 
epiphytes. From the fact of the species having such a local dis- 
tribution, it will probably be found that when the tropics have 
been thoroughly explored, that the number of species will be ma- 
