MR. F. N. WILLIAMS ON THE GENUS SILENE. 17 
And here, as in many other cases, it cannot be too clearly laid 
down that natural relationship is to be found in the possession 
of essential characters in the aggregate, and forms cannot be 
classified by the arbitrary selection of a single, albeit though a 
conspicuously distinctive character, and that the constancy of 
the specitic type is due to the maintenance of the balance and 
the absolute correlation of the structure of the different parts 
of the plant with the conditions of life and the environment 
in which it lives ; so that any undue exaltation or deficiency in 
the structure or function of any organ or part of an organ, 
whether arising within the plant itself or in its environment, 
tends to the impairment or to the loss of that balance, and 
consequently to instability and variation. 
Rohrbach first of all divides the section Botryosilene into 
two groups, according to the structure of the substance of the 
calyx, whether coriaceous or membranaceous; and then the 
second group lends itself more readily to division into sub- 
sidiary groups on the basis of the type of the form of racemose 
inflorescence. I wish to draw attention more particularly to 
this section rather than to any of the other primary groups, 
because if some working botanists hold that there are weak 
points in the method and arrangement in Rohrbach’s classifica- 
tion, it is in this section of Botryosilene that such are to be 
found. But, as he says, after a good deal of consideration and 
comparison of the relative value and specific importance of 
salient characters, he has not succeeded in finding a more 
natural method of associating the known species into definite 
subsidiary groups; and in any future improvement upon this, 
the critical reviser must have in his mind a general idea of the 
complete series of species, otherwise, instead of improving upon 
the scheme of classification submitted, he is likely to fall into 
worse errors. 
I will now, therefore, discuss a few specific characters in 
detail. A pre-eminent character to select is the form of the 
seed. Within the limits of the same group of allied forms the 
Seed varies chiefly in its size (generally depending relatively on 
the size of the flower and capsule), in its form and structure 
Scarcely at all. However, in a widely distributed and poly- 
morphous species like Silene inflata, we find transitional 
modifications in the form of the seed. On the other hand, 
almost the only difference between S. rigidula and S. echino- 
LINN. JOURN.— BOTANY, VOL. XXXII. Cc 
