BOTANICAL SUBJECTS. 207 



(c.) Free central placenta, that is, on a sort of axis, uncon- 

 nected with the carpels, except at their base, as in 

 Spergttlaria (Fig. 106). 

 This free central placenta has been much discussed. Some have 

 considered it the termination of the branch axis ; others as a 

 modification of parietal placentation with ovules produced only at 



Fig. 106. Spergularia. 



the bottom. But I think that in the stamina! tube of the Malvaceae, 

 we have perhaps the homologue of the central free placenta. 



I think there are two ways of looking at the origin of this 

 tube : — 



First. It may be formed by the fusion of the long filaments 

 of the stamens, the anthers remaining free. 



Second. The tube may be formed by fusion of the bodies of 

 appendages Hke those of the petal of Parnassia Caroliniana 

 (Fig. 107). These appear to be of the nature of branched 



Fig. 107. Parnassia Caroliniana (Le Maout and Decaisne, Fig. 4.57). 



glandular hairs, growing from the axillce of the petals in the same 

 way that a similar appendage grows from the axillae of the anthers 

 of Larrea Mexicana (Fig. 108). For we have no right to suppose 



