ON HYBRIDIZATION AMONGST VEGETABLES. 



91 



wet, and was also pestered by Aristolochia clematitis ; but G. 

 Illyrieus is only found in meadows of alluvial soil subject to 

 inundation ; and where I saw it in flower in May, near Trieste ; 

 the sod was then three inches under water. Yet these three 

 species require nice discrimination to separate them. Their case 

 is somewhat like that of the wrens, Silvia sylvicola, trochilus, 

 rufa, loquax, and Temminckiana, which are only distinguished by 

 skilful persons, yet differ greatly in note, nest, and habits. 

 Although the northern Gladioli, which conform with cardinalis, 

 &c. as to their seed, will not breed with them, I believe, on the 

 contrary, that there is no obstacle to their breeding with their 

 European congeners that have round, unwinged seeds. In the 

 genus Schizanthus, retusus refuses to breed with pinnatus; though 

 they conform with each other in all respects except the size of 

 their seeds — a circumstance which is not an obstacle in Gladiolus, 

 but is so in Anomatheca, as I have already stated. I can sug- 

 gest no direct solution for that mystery ; but some difference of 

 constitution probably prevents the pollen from deriving what is 

 essential from the juices of the female plant. I should conceive 

 that G. Byzantinus and communis, which have seeds like the 

 African sorts, and are not particular as to position, are nearest 

 to the northern Gladiolus of the oldest days ; that some of its 

 offspring, having fallen into peculiar situations, have acquired 

 constitutional peculiarities, with some alterations of aspect and 

 structure, that have become fixed characters. 



There are some classes of plants with great diversity of forms, 

 but so graduated as to render it almost impossible to subdivide 

 them satisfactorily into distinct genera and species. I find the 

 undescribed Colchicums from different localities varying so 

 much from each other generally, and yet so little fixed in their 

 own usual peculiarities, that all specific names for them seem 

 vain. Again, the difference which separates Colchicum from 

 Merendera, viz. a compact slender tube in the former, and in 

 the latter, instead of a tube, six long slender bases to the leaflets 

 of the limb, which are fastened together by little hooks at the 

 mouth of the seeming tube, but easily separable, would appear 

 to furnish a good generic distinction ; but the perfect agreement 

 of the two in habit and general appearance induces me to think 

 that they were united even in comparatively late periods of the 

 world, and I should expect to find them capable of intermixture, 

 and I shall take the first opportunity of making the experiment. 

 The Sisyrinchioid plants include a mass of individuals which it 

 is most difficult to class under distinct genera, but which are 

 very remote from each other in habits and appearance, and they 

 slide away through the Tigridialike plants towards the real Irises 

 and their near kin, in a manner that cannot be readily brought 



