268 PHILOSOPHICAL NOTES ON 



Fig. 108. Larrea Me.ricana (Asa Gray, "Stru. Bot.," Fig. 379). 



that because leaves have turned into petals and stamens, they are, 

 therefore, never capable of giving forth buds of some sort from 

 their axilla?. The appendages of Parnassia are probably groups 

 of abortive stamens,* the anthers being represented by the glands 

 at the tips of the hairs. 



Now, if in the axillfe of the carpels of Spergularia (Fig. 106) 

 similar appendages merged, and /w5<?c? into a co/«w7?, they would 

 produce the free central column of Primulacea3. In such a case, 

 the ovules would be transformed glands. The column would be 

 something like the staminal tube of the Malvacere. 



Asa Gray (" Structural Botany," p. 267) would explain the 

 central free placenta as a " parietal placentation with the wargins 

 of the carpels ovuliferous only at the Ijottom, and the placenta 

 there conspicuously developed, and completely united." This 

 mio-ht be so, but it appears to me that the axilla? of the carpels 

 would be quite equal to developing buds of the nature of those of 

 Parnassia, which by becoming united into a column would give 

 rise to the free central placenta of Spergularia. In such a case the 

 o\Tiles of this column would correspond to the glands on the teeth 

 of the leaflets, such as we see on the young stipules of the apricot 

 tree and other plants, the blades being contracted and fused into a 

 column. 



If, however, we take a broader view of the origin of phaenogams 

 in an evolutionary sense, we have only to turn to Trichomanes 

 radicans, Sw., to see, in the placentation of its sporanges, the 

 rudiments of the free central placental of phtenogams (Fig. 109). 

 There we shall see that the midrib of the cladophyl is continued as 



* They might be called either axillary stamens or axillar}- ovules. 



t According to Warming, the sporange was considered the homologue of 

 the ovule hy lirongniart, Cramer, Van Tieghem, and Celakovsky. He and 

 Gichler inclined to tlif same view. 



