228 



valve (bi) tends to disappear, it can, indeed, be the bearer 

 of ovules and even unité with the outer (bu) to a closed 

 whole. In this cases we may decidedly speak of a com- 

 pound pistil. 



It cannot, of course, occur to a botanist of the présent 

 day to infer from this fact that each stamen should be 

 comparable to two leaves, at best one could admit of a 

 splitting up of the connective in tangential direction. 



It we hâve succeeded in proving that in Nicotiana as 

 v^ell as in Sempervivum and Papaver the ovules hâve arisen 

 at both sides of the connective, the question romains 

 whether in botanical literature also other kindsof pistillody 

 of stamens hâve been registered. 



We leave, of course, the leap-variations unmentioned, 

 because the process of transformation défies observation. 



In the Mémoires de l'Académie de Toulouse, 5 Série, 

 T. III (1859) p. 105 D. Clos describes a tulip in which 

 one of the stamens was pistilloid, but in such a way that 

 the ovules arise between the cells of the anther, on the 

 connective. As figures are wanting it is better for a right 

 understanding to quote the exact words of the author: 



„En effet, au dessus du filet on distingue la base des 

 deux loges de l'anthère que surmontait une rangée d'ovules 

 imparfaits: ceux-ci naissaient à la jonction du filet et de la 

 loge anthérale avortée. Deux stigmates se trouvaient au 

 sommet de ce corps 



Cette observation démontre que les ovules n'occupent 

 pas la place du pollen, mais naissent à la jonction des 

 loges de l'anthère avec le connectif." 



The similarity to Nicotiana is in my opinion complète, 

 even where two stigmas occur. Both back and front of 

 the anther appear to hâve developed into a sort of carpel, 

 a fact also repeatedly found in Nicotiana. 



As an appendage to the interesting paper on the green 

 Rose (a monstrous variety of Eosa diversifolia) A Gris 



