92 



PKOSEKPINA. 



files, A and B, Fig. 11, are nearly straight-sided ; and 

 always, be it young or old, one broader than the other, so 

 as to give the flower, seen from above, the shape of a con- 

 tracted cross, or crosslet. 



6. Now I find no notice of this flower in Gerarde ; and 



B 



FIG. 11. 



in Sowerby, out of eighteen lines of closely printed de- 

 scriptive text, no notice of its crosslet form, while the petals 

 are only stated to be " roundish-concave," terms equally 

 applicable to at least one-half of all flower petals in the 



