MEMOIRS FROM THE DEPARTMENT OF BOTANY OF COLUMBIA UNIVERSITY 



7 



FLOWER-AXIS. 



The so-called calyx of the Rosaceae and other calycifioral families consists of two 

 distinct parts, viz., the upper free portion (the sepals) which constitutes the calyx as the 

 term is ordinarily understood, i. e., the outer set of floral envelopes, and a lower, more or 

 less cuplike portion. This has generally been regarded as the gamosepalous part of the 

 calyx, and is so described in almost all our American systematic works. It is, however, 

 without doubt of an entirely dijBferent origin. With the exception of, in some cases, per- 

 haps, a very small portion at the rim of the cup {i. e., just at the base of the free sepals), 

 the whole of the so-called gamosepalous part of the calyx must be regarded as belonging 

 to the flower-axis. This view is now held by most of the prominent systematists and 

 morphologists in Europe as Baillon,^ Henslow,^ Wiesner^ and most of the authors of 

 " Die Natuerlichen Pflanzenfamilien," edited by Engler and Prantl, viz., Engler,* 

 Focke,^ Gilg,® Raiman,' Koehne,* Schimper,^ Niedenzu,^" and Brandis." 



Some of the reasons for this view are as follows : 



1. The petals and stamens, which are generally regarded as homologues of foliage 

 leaves, are borne on the margins of the cup inside the calyx-lobes. If the cup is to be 

 regarded as the lower part of the united sepals, one set of phyllomes would be borne on 

 the upper surface of another set, which would be a very unnatural condition according 

 to our general conceptions. If the cup is regarded as a modified flower-axis, the ex- 

 planation of the arrangement of stamens, petals and sepals is very simple. 



2. If the cup is regarded as the result of the adnation of the lower portions of the 

 sepals, petals and stamens, the difficulty is not removed. The petals fall off" by a joint 

 at their insertion at the margin of the cup. This point should either be the lower end 

 of the petiole or else the petal should represent only the leaflet of a unifoliolate leaf. 

 In the flrst case it would be borne on the calyx as above. In the second case, the 

 difficulty would be that the leaflets in Potentilla and most genera are not deciduous, nor 

 even jointed to the rachis. We should also flnd a continuation of the leaf downward 

 representing the rachis. There is a kind of continuation in the form of a fibro-vascu- 

 lar bundle, but this unites with those from the antipetalous stamens before it reaches the 

 bottom of the cup. So also the tibro-vascular bundles of the sepals unite with those of 

 the antisepalous stamens. If the three sets of organs were completely united and all 



1 Histoire des Plantea, 1 : 345-6. 



2 The Origin of Floral Structures, 96. 



3 Elem. Wiss. Bot. 2 : 307. 



< Vol. 3 : part 2 a, 41 (Saxifragaceae). 



5 Vol. 3 : part 3, 4 (Rosaceae.) 



5 Vol. 3 : part 6 a, 100 (Loasaceae). 



'Vol. 3 : part 7, 199 (Onagraceae). 

 'Vol. 3 : part 7, 1 ( Ly thraceae) . 

 'Vol. 3 : part 7, 42 ( Rhizophoraceae) . 

 ">l. c. 57 (Myrtaceae). 

 "/. c. 106 (Combretaceae). 



