IN NATURAL CLASSIFICATION. 579 
in an apparatus; and likewise a constant relation between 
characters equally essential in different apparatusses in the 
same individual. 
This postulate is not difficult to admit; its truth affords one of the 
principal charms to physiology. Thus, the nature of the inflorescence regu- 
lates the form of the corolla; when the one is terminal, or furnished with 
long and remote peduncles, so that the flower is equally preset. on and 
equally nourished, the corolla is regular; whereas, if t be 
athickly congested spike, umbel, or capitulum, the part most vi on 
and least nourished, is least developed, and the flower is irreg Eve 
when the connection of cause and effect is not perceptible, Ce SE of 
the conditions of existence is not the less marked; thus, in flowers truly 
unisexual, the perianth becomes masked; in this case, although the proxi- 
mate cause is concealed, the final cause is evident, inasmuch as the pre- 
fecundation. Again, in many instances, we observe a constant co-exist- 
ence of characters, where our imperfect knowledge ate insufficient to reveal 
to us either the proximate or final cause of this coincidence. Thus stomata 
and spiral vessels always co-exist, but wherefore we do not know 
This postulate, however, not only requires us to admit a constant rela- 
tion between characters of unequal importance in an apparatus, but also 
between characters of tnl importance in different — Now, 
Observation likewise proves the truth of this position; for do we not see 
e nearly constant co-existence of a monocotyledonous Pena with an 
endogenous stem; a dicotyledonous embryo with an exogenous stem; 
naked ovules, with the elongated ellise es studded with glandular disks ? 
—and I might quote many coincidences of a like nature. 
Tf the foregoing postulates be admitted, we can found on 
them the following propositions : ' 
lst.—A reunion of similar apparatusses agreeing in the most 
essential characters, is a reunion ofapparatusses having a 
Constant relation in less essential characters. 
For we have already granted that an apparatus is a collection of organs 
engaged in the performance of the same function; therefore the appara- 
tusses of different € that are for the execution of the same func- 
tion in each is similar in that respect; and we have also admitted that 
Some characters are more, and some less important to the performance 
of a cernes = that the organs engaged in fr e performance of a func- 
nar ; and that the conditions 
