GLADIOLUS ILLYRICUS AND ITS ALLIES. 133 
viduals upwards. A Natural Order would be very ill defined by a 
botanist who was ignorant of the characters of the genera which he in- 
cluded in it, or a genus by one unacquainted with its component 
species. No doubt this is frequently done, and then some more care- 
ful observer comes into the field and upsets the work of the first. In 
the same way, superspecies can only be satisfactorily established by 
those who are acquainted with the most restricted groups of permanent 
forms which compose them. It is, in a great measure, owing to this, 
that botanical nomenclature is so fluctuating. The imperfect know- 
ledge which the founder of a species has of the plants which he includes 
in that species, often leads him to admit under it aberrant forms be- 
longing to some other type, and to exclude aberrant forms of his own 
species. The only real starting-points are individual plants, among 
Which we find some forms which are héreditarily constant in those 
points in which they differ from others, producing races which are 
practically permanent for such periods of time as our observation ex- 
tends over. Some of these races are comparatively widely separated, 
even from those which most nearly resemble them, and such constitute 
what Mr. Watson has termed verspecies, about which there is no 
difference of opinion. Others, again, approach much more nearly their 
neighbouring races; these are the so-called subspecies, which have to 
be grouped together to compose superspecies. But if superspecies 
are formed without an accurate knowledge of their subspecies, artificial 
instead of natural groups are likely to be the result. 
pene number of European species of Gladiolus enumerated in Koch's 
“Synopsis,” Ledebour’s ‘Flora Rossica, Grenier and Grodron’s ‘ Flore 
de Frauce, Parlatore’s ‘ Flora Italiana,’ and Willkomm and Lange’s 
"Prodromus Flore Hispanics,’ is twelve, of seven of which I possess 
specimens, These seven appear to belong to three species, two of 
which are superspecies :— 
1. G. communis, Liw., including C. communis, Koch (eu-commu- 
his mihi), G. I/]yricus,* Koch, and probably also G. imbricatus, T auct. 
plur. (uot Linnwus), and G. palustris, Gaud. 
Eoi ll, s 3 H ; 
- HP ‘ spicatus of Linn. Herb. ! ted by a deformed spe- 
cimen of G. cu-communis !, with the flowers crowded together, from the spike not 
having lengthened before the flowers expand. 
LA The, meshes of the fibrous covering of G. J/yrieus from Lyndharst are occa- 
Sionally quite as wide as those of G. palustris. 
