so far as I know, in a wild state it varies little in colour. The 

 variously tinted garden forms of which I spoke just now, the red- 

 purple and the almost red kinds, as well as the very common 

 forms in which a white ground is more or less splashed with 

 blue or blue-purple or red-purple blotches, are the outcome of 

 the repeated seeding to which this species for some two or 

 three hundred years has been subjected in cultivation. But 

 as I said, a yellow plant is unknown ; this colour has never 

 made its appearance during the many, many generations of 

 seedlings. Moreover, so far as I can ascertain, though repeated 

 seeding has produced great variety of colour, it has hardly 

 affected at all the structural characters of the plant ; the 

 various forms now cultivated, apart from size and colour, are all 

 exceedingly alike. This is interesting in connection with the 

 narrow geographical distribution of the species. Iris xiphioides, 

 indeed, may be regarded as the type of a really good species. It 

 differs from all its allies by characters so broad as to be obvious 

 to everyone ; it exhibits little or no tendency to vary, or to form 

 hybrids with other species. It at some time or other acquired 

 certain features, and those early became so rigidly fixed that it 

 speedily lost all power of adapting itself to varied circumstances, 

 and hence has proved unable to spread outside a very limited home. 

 The Spanish Iris, I. xiphium, on the other hand, has not 

 only a much wider range, spreading throughout the greater part 

 of Spain and Portugal into the African continent, and reaching 

 both into France and into Italy, but also comes very close to 

 other species ; so much so that between them and it the ques- 

 tion of specific differences is soon raised. Among the wild forms 

 two types may be recognised. In the one, the falls, which 

 are relatively narrow, spread out horizontally, the ovary pro- 

 trudes from the spathe-valves for some distance, and the prevail- 

 ing colour is blue or purple. In the other, the claws of the 

 falls, which are relatively very broad, rise up in a slanting 

 fashion, so as to form more or less of a funnel ; the flower 

 is " turbinate " ; the ovary is much less exserted, and the 

 dominant colour is yellow. The latter form is found in Portugal, 

 and hence, though the other blue form is also common in that 

 country, has been called I. lusitanica ; a variety of it, in which 

 the colour is not pure yellow, but heavily blotched with brown 

 is the I. sordida of Salisbury. 



