552 Mr. MiEiis o)i some new Brazilian Plants 



placentae, a relation most usual in cases of compound unilocular ovaria where 

 tile number of stigmata and placentae is equal ; and that such is really its 

 relation appears to him to be proved by tracing to their origin their vascular 

 cords, which are found to coalesce M'ith those of the three outer foliola of the 

 perianthium. This view of the composition of the ovarium in Orchideœ, he 

 observes, is confirmed by finding that it agrees with the ordinary arrangement 

 of Monocotyledonous plants, viz. the opposition of the double parietal pla- 

 centœ to the three inner divisions of the perianthium, while in Apostasia the 

 three placentre of the trilocular ovarium are opposite to the three outer divi- 

 sions. The same agreement, he further observes, is found in Scitaminecv, both 

 in the placentee of the trilocular ovarium, which in this family is its ordinary 

 structure, and in the unilocular, which is the exception. My observations 

 upon the structure of Burmanniaceœ alïbrd to that order a different arrange- 

 ment as regards the position of stigmata. Di'. Von Martius, in illustrating 

 the genus Bumunmia, has given a figure of the pistillum of B. bicolor, in 

 which the stigmata are placed opposite to the wings, and therefore alternate 

 with the inner segments of the perianthium ; but this probably may have been 

 an error of the draugiitsman, since no such position is alluded to in the text. 

 I have in several instances opened with the utmost care the flowers of Bur- 

 mannia, and have found the stigmata manifestly placed as I have constantly 

 observed them in Dietyostega and the allied genera with unilocular capsules, 

 viz. opposite to the stamens, and to the inner segments of the perianthium, with 

 which the placentee also correspond, all being alternate witli the outer seg- 

 ments : in Burmannia, however, owing to the complete inflection of the car- 

 pellary leaves to form the trilocular ovarium, the placentœ thus extended 

 to the axis will be seen directed towards the middle of the cells, and opposite 

 to the outer segments of the perianthium, at the same time that all other parts 

 remain as before mentioned, in a similar position to that existing in Die- 

 tyostega. 



This deviation from the usual order of relation may probably be accounted 

 for by the very ingenious views of Mr. Brovrn i-elative to the original compo- 

 sition of stigma, founded on the supposition that each simple pistillum or 

 carpel has necessarily two stigmata, which are to be regarded not as terminal, 

 but lateral, in the same manner that the placentœ of each carpellary leaf are 



