NUMERICAL I5CREASE OF PARTS OF PLANTS. 209 



Secondlj, when flowers become double tliey may lose tlieir ver- 

 ticillate character and become spiral. Thus, I counted more than 

 forty petals in a double Wallflower. 



4. Pleiophylly . — As in the ease of polyphylly, the abnormal in- 

 crease of the number of leaflets of foliar appendages may be either 

 due to an accidental numerical excess, or to atavism, one or more 

 leaflets being usually suppressed. Thus, compound leaves occa- 

 sionally develop extra leaflets, as the 4-7-leafed Shamrock (TV/- 

 ^ folium repe7is) ; so also does OsaJis, Laburnum, &c. ; while the 



AVillow called Salix pendula multistipulata develops several sti- 

 pules. Such abnormal varieties may become permanent. 



Some leaves, however, such as of the Elm, occasionally produce 

 a supernumerary leaflet at the base of the blade (Z, c. p. 353, fig. 

 183). This is a restoration ; for the leaf is normally oblique in 

 consequence of the arrest of development of the lowermost fibro- 

 vascular bundle on one side. If this be restored it produces a 

 small leaflet occupying the space usually left vacant. A still 

 further excess of development blends the leaflet into the main 

 portion, and the leaf becomes simple and no longer oblique*. 

 ^ III. Supplemental Deyelopments. — Under this head I would 



include all processes of enation {I, c. p. 443) — that is, outgrowths 

 from organs after they have emerged from the receptacle or axis ; 

 such are spurs to corollas, which arise as saccate bulgings which 

 become prolonged into spurs. Some forms of such processes, 

 however, are not thus formed by the entire structure, but only by 

 ^fold in one epidermis. To such I apply the word plication. The 

 Tudimentary form maybe seen in the " corona " of Forget-me-not, 

 in Primula^ and, in a more exaggerated condition, in tliat of Nar- 

 cissus. In the latter example fibro-vascular bundles pass up the 

 innermost and outermost membranes, but none occur on the outer 

 epidermis of the corona, or on the inner or upper epidermis of the 



perianth-leaf. 



IV. Chorisis, Fission, or Bifubcation. — The causes of bifur- 

 cation are not clear. In some monstrous instances it may be 

 simply due to injury at the apex of the organ. It may occur in 

 axes, as when two cones or catkins take the place of one, or when 

 the stems of ferns and the roots of Lycopodium bifurcate. Again 



* The cause of the obliquity of the Elm, Lime, &c., is, I believe, due to pe- 

 culiar vernation, as I have elsewhere stated (Linn. Trans, vol. i. part 2, p. 37, 



The stipules lie obliquely across the eJges of the conduplicate leaf, 





•and thus allow one side to develop more readily than the other. 



