118 TAXONOMY. 



By studying these relations, by carefully comparing the 

 structure with the function, we arrive at a sure knowledge of 

 the nature of organs. 



176. 



Mistakes in this respect are committed, when we overlook 

 three things, namely, the Abortion, Alteration, and Union of 

 Organs. Respecting all these three phenomena of nature 

 we must now give a more particular account. 



A. Of Abortive Organs. 



111. 



The abortive state of an organ is often the consequence of 

 an imperfect evolution. The cause lies not unfrequently in 

 unfavourable weather, in an unfertile soil, and in other acci- 

 dents. We thus see that fruit is not completely formed, 

 when the soil is arid. But the circumstance of most import- 

 ance is, that abortion arises very often from a law of Na- 

 ture, in consequence of which one part is evolved at the ex- 

 pence of another ; and this latter part, therefore, remains im- 

 perfect. We every day see a remarkable example of this in 

 many species of Violets, where some blossoms, in which the 

 corolla is chiefly unfolded, fall off, without leaving any fruit, 

 whilst other flowers, which had not a corolla, set perfect fruit 

 The beautiful colouring of the bracteae of Buginvillaa, leaves 

 no distinct evolution to the corolla ; and in our Melampyrum 

 nemorosum and cristatum, the flowers do not expand on the 

 upper parts, where the beautii'ully coloured bracteae seem to 

 supply their place. It is a law of Nature, which remains 

 constant even in entire families, that half of the filaments 

 are abortive, or that the loculi of the fruit, originally des- 

 tined undoubtedly for the reception of seed, remain partly 

 empty. In the Acantheae and Sapotea% we find the pheno- 

 mena of abortive filaments as a law of these families. In the 

 Rhamneae, Palmae, Sapindeac, and many other families, the 

 abortion of the individual loculi of the fruit is constant ; 

 and who can have compared our Gaura biennis, in the 



