I 



AHBANGEMENT OF THE PARTS OF THE FL0WEB. 475 



by descriptive botanists, whose attention has not been specially 

 called to the question of relative position. 



For the purposes of this communication it matters relatively 

 little whether the genera called Olacaceous are closely allied or 

 not in their genetic affinities, as abstract relations of position are 

 the main subjects now under consideration. This being the case, 

 it is not necessary in the formula? made use of to indicate any dif- 

 ferences of adhesion, cohesion, or (except in some cases) number of 

 parts. The position of the loculi of the ovary is also, in general, not 

 indicated — because in most instances there is only a single cavity 

 in the adult flower, or, if there be more than one, then it is difficult, 

 if not impossible, among Olacaceae, accurately to determine the 

 relative position of the ovarian cavities from dried specimens. In 

 Emmotum, according to Miers, there are three ovarian cavities, with 

 indications of two others. 

 The simplest formula for this order is 



S5 



St 5, 



which is met with in Lepionuras, where we have a calyx of five 



sepals with five stamens superposed. That the superposition in 



this genus is due to suppression of the petals may reasonably be 



inferred from the fact that one of the most common of all floral 



arrangements is 



S 

 P 



St; 

 and this is found in many Olacaceous genera, such as Mappia, 

 Lmianthera, Gampliandra, Cardiopteris, Apodytes, Pteleocarpa, 

 Phlebocalymna, Sarcostigma, Phytocrene, Miquelia, &c. 



The arrangement 



S5 



P5 



St 5 



m>fi 



<Anacalo8a, and may, possibly, be accounted for by reason of the 

 absence or suppression of an outer antisepalous series of stamens. 

 In Ojnlia the arrangement is similar ; but there is in this genus 

 a row of staminodes superposed to the sepals and within the 

 stamens 



