or saponabia officinalis. 161 



adherent to the claws of the petals. There were also numerous 

 cases showing an intermediate condition between the ordinary- 

 scale and that of two distinct stamens — some of these are drawn 

 in the sketch given below. A careful inspection of some very- 

 young flower-buds showed that these scale-like stamens, though 

 formed subsequently to the petals,, yet advanced more rapidly in 

 their development than the petals, from which in the first instance 

 they were perfectly distinct. 



From these circumstances it seems fair to infer that the scales 

 on the petals of the CaryopJiyllece are composed of two abortive 

 stamens united together, and in consequence that these bodies do 

 not constitute any real exception to the laws of alternation. 



The nature of the scales of the flowers of the species of Cuscuta 

 has been explained by Mr. Babington in the Annals of Nat. Hist, 

 for 1844 in a similar way, but he does not adduce any positive 

 evidence in support of his opinion. 



The flowers of Saponaria above alluded to afforded instances of 

 both marginal and free central placentation. In the sketch are re- 

 presented cross sections of two ovaries (figs. 1,2), each consisting 



of two carpels, and in each there is an evident connexion between 

 the placenta and the walls of the ovary. Other bi-carpellary ovaria, 

 examined at a more advanced period of development, presented a 

 central ovuliferous column, apparently quite unconnected with 

 the walls of the carpels. The most frequent condition of the ovary 

 was that in which there were three carpels, slightly united at their 

 bases, but distinct one from the other for the greater part of their 

 length, open along the ventral suture, and bearing ovules on their 

 margins, as is shown in transverse section in the sketch (fig. 3). 

 Figure 4 shows one of three carpels ; the other two bore marginal 

 placentae, but in this one there was an ovuliferous cord, free from 

 the margins of the carpel, but connected with the style, which is 

 inflexed. Here it seems as if the two placentae had become 

 detached from the margins of the carpel during growth, while 

 they had remained in union one with the other. 



A not unfrequent condition was that in which there were two 



LINN. PBGC. — BOTANY. M 



