NATURAL CLASSIFICATION. 115 



parts of the seed. 3. The fruit. 4. The filaments and an- 

 thers, the latter of which must be more important than the 

 former, because the former exist but for them. 5. The nec- 

 taries, which, when they are present, essentially promote 

 fructification. 6. The interior cover of the sexual parts, or 

 the corolla. 7- The calyx, or exterior cover. 



172. 

 There is a second mean of determining the value of or- 

 gans, which, however, although very instructive in certain re- 

 spects, is liable to more objections. We may consider any 

 part of fructification as having so much the greater import- 

 ance, according to the number of species in which it is found. 

 By following this rule, we obtain nearly the same results as 

 from the former ; because the object of vegetation is in all 

 cases affected, when the same means are not in operation. The 

 seed, or germ-grain, is universally present, even in the lowest 

 plants. The individual parts of the seed cannot be so com- 

 monly distinguished. All plants have not fruit; but more 

 shew fruit than sexual organs. Of these organs, the female 

 parts, even those which are shortest lived, are found in more 

 plants than the male parts; since even in the Homalo- 

 phyllae, and Musci hepatici, there are pistils and stigmata, 

 where no anthers have yet been shewn. Anthers occur more 

 commonly than filaments. We dare not decide whether nec- 

 taries or corollge appear most frequently ; since as many 

 flowers wanting the corolla have nectaries, as there are nec- 

 taries without the former organ. As little can we affirm that 

 the corolla appears more frequently than the calyx, because, 

 in numberless cases these integuments pass into each other. 



173. 



A third mean of judging respecting the importance of or- 

 gans, consists in observing how far a certain oigan is more or 

 less constantly united with the structure of definite and gene- 

 rally received vegetable tribes. If, for example, we had to 

 determine whether the stipulse or spines are most im})ort- 

 ant, we must give the preference to the former, because there 



II '2 



