' . inated in variation) would seem to indicate t 
J. D. Hooker, Introductory Essay to the Flora of Tasmania. 17 
hold its own, and not succumb to the enervating or etiolating 
or smothering influences of its neighbors. The effects of this 
warfare are to extinguish some species, to spare only the hardier 
races of others, and especially to limit the remainder both as to 
area and characters. Exceptions occur in plants suited to very 
limited or abnormal conditions, such as desert plants, the chief 
obstacles to whose multiplication are such inorganic and princi- 
pally atmospheric causes as other plants cannot overcome at all; 
such plants have no competitors, are generally widely distrib- 
uted, and not very variable. 
15. The three great classes of plants, Acotyledons, Monoco- 
tyledons, and Dicotyledons (Gymnospermous and Angiosperm- 
ous), are distributed with tolerable equality over the surface o 
the globe, inasmuch as we cannot indicate any of the six con- 
other. Further, the distribution of some of the larger orders is 
remarkably equable, as Composite, Leguminose, Graminec, and 
others; facts which (supposing existing ge to have orig- 
t : 
which is analogous to that already stated (4), that the least com- 
_ plex are also the most variable. 
* Though invariable forms, they may be, and often are, themselves Mgnt 
and constant form of P. pratensis, occurring in dry sandy soil, from 
+t Very much, no doubt, because of the difficulty in classifying Dicotyledons by 
complexity of organization; in other words, of our inability to estimate in a classi- 
tory point of view the relative value of the presence or absence of organs in 
plants, where many are present, and where those of low morphological importance 
May have a comparatively high physiological significance. 
SECOND SERIES, Vou. XXIX, No. 85.—JAN., 1860. 
3 e 
