168 PRINCIPLES OF PLANT-'I'ERATOLOGT. 



Bearing this theory in mind, and also the view, above 

 set forth, that the ventral lamellae of the virescent 

 anther and of the abnormal leaves of saxifrage, etc., 

 represent an imperfectly formed separate leaf,* it is 

 interesting to note what a strong resemblance to a dis- 

 tinct and independent foliar organ the fertile " spike " 

 of the Ophioglossace^ bears. 



Morphological Structures of the Lower Surface. 



There remains yet to explain, along the morphological 

 lines of anther-formation, those cases in which two 

 foliage-leaves become fused, or separated off, along the 

 midribs of their lower surfaces, as in the mango- and 

 tobacco-leaves previously described, and the occurrence 

 of enations and pockets on the lower surface, as in 

 Polygonum cuspidatiim,cab\)age, Ficus be7ighalensis, etcf 

 It will now be shown how anther-formation may quite 

 well have arisen from such structures also; this can be 

 understood much better by a reference to the figures 

 than from any description. We have to imagine the 

 four lamellae, which arise from the dorsal fusion, or 

 separation off, of two leaves, becoming inroUed in pairs 

 to form four loculi with the upper surfaces innermost. 

 By comparing figs. 4 and 5, PL XLIV, we can further 

 see that the two loculi formed by the inroUing of the 

 lamellae of each original leaf (me and ma, fig. 128) cor- 

 respond respectively to a median and marginal loculus 

 which are formed, a pair of such on each side, by the 



* I. e. when dichotomy or fusion occurs in the plane of the leaf -surface. 



t In the aroid Xanthosoma atrovirens there has appeared a sport known as 

 Caladium appendiculatum X in which all the leaves during the plant's life- 

 time exhibit lamellar enations from the midrib of the lower surface. These 

 may be simple, or may themselves produce similar enations, but, according 

 to Andre's account, the secondary ones do not obey the law of laminar 

 inversion, as do the primary ones ; or, again, there may occur two sets or 

 pairs of lamellae whose ends may become united to form semi-ascidial 

 structures, the morphologically upper surface of each being innermost. 

 Here then we see transitional structures between asoidia and ventral 

 lamellEe which tends to support the view that the meaning of both is the 

 same. 



X Herbarium specimens of this sport were seen at Kew. 



