l(i 



DIMORPHIC LEAVES IN RELATION TO HEREDITY. 



(liniorpliic variety than in the lobed leaves. The smallest teeth are 

 found on the specimens with the very nan'ow digitate lobes. (See 

 figs. 5 and 6.) 



Examples of transition forms of leaves seem to be more common 

 on plants with rather small, narrow-pointed, sharjoly dentate leaves 

 than in plants with larger leaves and less numerous teeth. (See 

 PI. II, A, B, C, and D.) It is not impossible that these differences 

 represent distinct varieties or strains. There is no reason to suppose 

 that the Egyptian varieties of this plant have been subjected to any 



more close or careful selection than the Egyp- 

 tian varieties of cotton, which were found to 

 exhibit a wide range of diversity. 



In Hooker's "Flora of British India" the 

 leaves of Hibiscus cannabinus are described 

 in two slightly different ways, once "Lower 

 leaves entire, upper lobed," and again "Lower 

 leaves cordate, upper deeply palmately lobed, 

 lobes narrow serrate." The narrow-lobed 

 variety shown in figures 5 and 6 would seem 

 to conform most nearly to this description, 

 though none of the lower leaves are shown in 

 the pressed specimens of tliis variety in the 

 Economic Herbarium of the United States 

 Department of Agriculture. The species 

 seems not to be represented in the National 

 Herbarium. The leaves of the Egyptian 

 varieties would hardly be described as cor- 

 date, though some of those in Plate II show 

 a slight reentrant angle at the base. 



The lobing of the leaves of the dimorpliic 

 Egyptian variety is not unlike that of the 

 plant depicted as Ilibiscus cannabinus in Rox- 

 burgh's "Plants of the Coast of Coromandel" 

 (vol. 2, pi. 190), except that some of the upper 

 leaves are shown with five lobes. Though no 

 such leaves were seen on the Egyptian plants 

 in July it is quite possible that they occur later in the season. Rox- 

 burgh also gives a separate figure of a simple narrowly lanceolate leaf 

 and states that this form occurs at the toj) of tlie fidl-grown plants. 

 According to Wester a similar reduction of the later leaves is shown 

 in the roselle plant {Hibiscus sabdariffa)} 



• Wester, P.J. Roselle: Its Culture and U>ses. Farmers' Bulletin 307, U.S. Dept. of Agriculture, 

 1907, p. 7. " The leaves on the young plants are entire; as the plant increases in size the loaves change 

 to palmately five parted; later the leaves in whose axils the flowers are borne are three parted." 



221 



Fig. 5.— Three-lobed leaf of nur- 

 row-lobod variety of Ilibincus 

 cannabinus, grown in Louisi- 

 ana. (Natural size.) 



