132 Development of the Natural System under [BOOKI. 



the abortive leaves of the vetch and the abortive inflorescences 

 of the vine are employed as tendrils. In other cases the 

 abortive organ appears to be quite useless, as for instance 

 many rudimentary leaves. All such useless organs, says De 

 Candolle, exist only in consequence of the primitive symmetry 

 of all organs. Finally the abortion may be so complete that 

 no trace of the organ remains, of which case there are 

 however two kinds, one where the organ is at first perceptible 

 and afterwards quite disappears, as in the abortive loculaments 

 in the fruit of the oak ; in other instances no trace is to be 

 seen from the first of the abortive organs, as happens with the 

 fifth stamen of Antirrhinum. 



All that has here been said might be alleged word for word 

 in proof of the theory of descent, but our author is an adherent 

 of the dogma of the constancy of species ; what from his point 

 of view he really means by abortion is difficult to say, for the 

 object which is aborted is wanting. If species are constant, 

 and therefore of absolutely distinct origin, we must not speak 

 of abortion ; we can only say that an organ which is present 

 or large in one species is small or wanting in another. In 

 introducing the idea of abortion De Candolle at once goes 

 beyond the dogma of the constancy of species, without being 

 clear in his own mind with regard to this important step. His 

 proceeding shows that facts lead even a defender of constancy 

 against his will to theories which run counter to that dogma. 

 This is confirmed by his perception of the correlation of 

 growth, which is connected with abortion ; he points to the 

 fact that owing to the disappearance of sexual organs in the 

 disk-flowers of Viburnum Opulus the corollas become larger, 

 as do the bracts of the abortive flowers of Salvia Horminum ; 

 similarly he regards the disappearance of the seeds in Ananas, 

 Banana, and the Bread-fruit tree as the cause of the enlarge- 

 ment of the pericarps ; it does not escape him, that the fertile 

 flowerstalks in Rhus Cotinus remain naked, while an elegant 

 pubescence forms on the barren ones ; the leaf-like expansion 



