50 On the Production of Hybrid Vegetables. 



having done so. Its pollen, examined in the microscope, 

 does not shew much appearance of being prolific. In the 

 germen of one of its flowers, although two of the cells were 

 completely abortive, I did find one ovulum tolerably per- 

 fect, but ill-formed, and it is therefore possible that it may 

 occasionally produce a seed ; but I am confirmed in a sus- 

 picion which I entertained on the first sight of its inflores- 

 cence, that it is a natural mule produced in Sumatra, between 

 a campanulate and a radiate Crinum : I should have said, 

 without hesitation, between Crinum toxicarium or C. Suma- 

 tranum and Crinum Zeylanicum, (especially as it shews 

 faintly the red point to the youngest leaf, which distinguishes 

 C. Zeylanicum,) if its flowers had not been longer and of a 

 deeper purple, than such an union would appear likely to 

 produce; but perhaps the size and colour of its flowers may 

 have been occasioned by some favourable circumstances of 

 soil and climate ; or the male parent may be an unknown 

 species, superior in those points to C. Zeylanicum. I 

 also strongly suspect that Crinum longiflorum,* (which is 

 Amaryllis longifolia var. longiflora of the Botanical Regis- 

 ter, Plate 303), is also a hybrid plant, produced from C. Ca- 

 pcnse, which, I believe, is naturalized in Demerara, as the 

 Cape-coast sorts are near Rio Janeiro ; in leaf and stem, it is 

 scarcely distinguishable from the seedling mules between C. 

 erubescens and C. Capense. I should have very little doubt 

 of producing a plant very like C. amabile, though with shorter 

 and paler flowers, by impregnating any of the tall columnar 

 white Crinums with the dust of C. Zeylanicum, but I think 

 it would be sterile. 



* See the Enumeratio Specierum of the genus Crinum, Potanical Magazine, 2121. 



