234 PRINCIPLES OP PLANT-TERATOLOGY. 



the petiole, of any adnation of an axillary inflorescence 

 with the leaf, he declines to accept this explanation. 

 Yet it must surely be regarded as a case of adnation, 

 for adventitious flowers as "enations" from a leaf- 

 surface are entirely unknown. Moreover, the author 

 is mistaken in supposing that the anatomical structure 

 of the petiole can be a reliable guide, one way or 

 the other, in helping to solve such a morphological 

 problem. There are several normal instances of 

 "epiphyllar" flowers, such as those of Phyllonoma 

 ruscifolla (Saxifragacese) and Erythrochiton hypophyllum 

 (Rutacese), in which a similar appearance is presented 

 to that of the Amaranthus, and in which the adnation 

 is very intimate and complete. It is possible that the 

 union in Amaranthus might be so intimate that no trace 

 of it, external or internal, could any longer be found. 



Excellent normal examples of the fusion of an 

 inflorescence-axis with its bract are afforded by the 

 lime [Tilia) and by Spathicarpa (Aroide^). 



We come now to the adnation of floral leaves. 

 Moquin-Tandon cites a case of the adnation of sepals 

 to petals in Geranium nodosum in which a petal clave 

 by its lower surface to one of the calyx-segments. 



There were cited on a previous page instances, viz. in 

 those flowerswhere the cyclic tended to become replaced 

 by the acyclic arrangement, of a sepal becoming fused 

 with a petal laterally, so that the double perianth-leaf 

 so constituted belonged partly to the outer and partly 

 to the inner whorl, and was also of dual nature ; e. g. 

 in the crocus, tulip, and lily. Buchenau mentions a 

 similar instance in Tropseolnm majus. 



A drawing, kindly lent by the late Mr. Gr. S. Saunders, 

 showed an interesting case of a flower of Gypripediuvi 

 superhiens (PI. L, fig. 1) in which the two lateral 

 petals were fused by their margins with the posterior 

 sepal (i. e. the one behind the labellum) to form a 

 single large leaf ; the labellum, in correlation with this 

 change, being quite absent (fig. 141). PI. L, fig. ,2,* 



* Also lent by Mr. Saunders. 



