6 MR. F. N. WILLIAMS ON THE GENUS SILENR., 
The characters which are thus selected for the purpose of 
grouping allied genera are of primary importance; to these 
may be added two genera widely diverse from one another, but 
which are apparently of lower grade in differentiation, and the 
species of which by some systematic botanists would be very 
properly distributed among one or more of the other genera. 
These are—FPetrocoptis and Heliosperma. The first is evidently 
very near Coronaria; the species of the other combine the 
unilocular capsule of Melandrywm with the habit of Silene. 
The salient character of both is found in the morphological 
character of the seeds, which after all is of specific rather than 
of generic value. Though the presence of an appendix or ligule 
at the junction of the claw with the blade of the petal is of 
subordinate specific importance, we find that the petals are 
imbricate (in the narrow sense) in prefloration in the case of 
Petrocoptis, while in Coronaria they are strictly convolute in 
preefloration. 
In the delimitation of genera and transference of groups of 
species which have taken place from time to time in the suborder 
Sileninez, no genus has probably received such rough handling 
and mutilation, more particularly at the hands of critical 
systematists in continental floras, as the genus Lychnis. Even 
in an attenuated Linnean sense it is not so much as admitted 
into some of the German floras; while the compilers of various 
English floras, rather than introduce strange names into their 
lists of genera, have indefensibly enriched Szlene at its expense. 
The only absolute difference between Silene and Lychnis, as 
defined by Linneus, was that the former had three styles, and 
the latter five; and Agrostemma is only distinguished from 
Lychnis in having the lamina of the petal undivided. However, 
as the species of these three genera came to be more carefully 
studied, it was soon apparent that they should either be fused in 
one genus and broken up into natural sections, or that new 
genera should be formed out of them, in which the number of 
the styles should be considered as a character of very secondary 
importance, and in which the general structure of the ovary 
and capsule should determine the grouping of the species. 
A unilocular capsule, occasionally plurilocular at the base, is 
characteristic of the natural order Caryophyllacew; and this 
character has been selected for grouping the species of Silenoides 
(after eliminating the genus Cucubalus) into two other primary 
