REJUVENESCENCE IN NATURE. 97 



abnormal transitions between these two formations, a 

 transformation * of stamens into carpels, or, retrogres- 

 sively, of carpels into stamens,! while the transitions 

 between calyx and corolla, as well as between petal- and 

 stamen-formation are comparatively far more frequent. 

 The abortive circles must here be accounted partly to 

 one and partly to the other of the formations adjoining, 

 the altered conditions of arrangement and number so 

 frequently occurring in the fruit affording a distinct sup- 

 port for this in many cases. The abortion of an inner 

 circle of stamens is exhibited most convincingly by many 

 species of Jmicus, which occur sometimes hexandrous and 



* I use this term in the sense explained above, pp. 60, 61. 



\ The occurrence of such transitions is in most cases a certain indication 

 that no abortive circle exists, but the cases must be closely examined, in 

 order that a mere multiplication of the organs of one or other formation, 

 such as may happen through an alteration of relative position or through 

 axillary formations, may not be taken for a substitutive aberration of 

 structure. Consulting the sections on the transformation of pistils into 

 stamens, (218), and stamens into pistils, (220), in Moquin Tandons 

 ' Teratologie,' we find that after separating the doubtful from the trust- 

 worthy and accurately known cases, tlie examples mentioned belong to 

 three families of the Monocotyledons, (the Liliacese, Colchicacese, and Palmae), 

 and eight families of Dicotyledons, (the Ranunculaoese, Magnoliaceaj, Papa- 

 veraoese, Cruciferse, Crassulacese, Ericaoese, and Primulaceae). I have myself 

 observed most of these, as well as many other less known cases. To the 

 latter belong, for instance, the transformation of the carpels into stamens in 

 Alliiim Soheenoprasum, which seems to occur as commonly in the chives 

 cultivated in gardens, as the opposite case of the inner stamens turning into 

 carpels in the cultivated Sempervivum tectorum,. This transformation is 

 exhibited in the chive in the most varied degrees ; it is further remarkable 

 from the fact that the stamens appearing in the place of the three carpels, 

 have extrorse anthers, while the anthers of the six normal stamens are 

 introrse, a condition which reminds us of the similar double character of the 

 anthers in the Laurineai and Polygoneae. Strange too, is the occurrence of 

 three more shorter stamens, which are confluent with the three replacing the 

 carpels, and which I can only regard as axillary structures like those 

 occurring in double flowers, (see note, p. 79.) In this case we find inside 

 the decomposed germen, which has passed over into a staminal formation, a 

 new more or less perfectly developed whorl, the organs of which alternate 

 with those of the preceding whorl. Another case, apparently also little 

 known, but equally frequent, occurs in the cultivated horse-radish {Armoracia 

 rmticana). The two carpels are here transformed more or less completely 

 into stamens, while two other organs, absent in normal flowers, make their 

 appearance as carpels. The reverse ease, the transformation of all the 

 stamens into carpels, is shown by Cheiranihm Gheiri gynaniherus, Dec, a 

 monster which has become a variety in the Paris garden, and for which I 

 am indebted to M. Gay. 



7 



