8 
E. J. Salisbury. 
that the centre of variation is not single but multiple; even those 
forms which we do, after careful consideration, group together as 
continuous variations may, after all, be but an aggregate of strains, 
the differences between which are so small as to give the series an 
appearance of continuity. Recent research seems to point more 
and more to the external and visible differences as criteria of change 
established, relegating to the internal and invisible the steps which 
lead up to these mutations. 
In the case under consideration there appear to be several well 
defined types around which variations of a lesser degree may be 
grouped. But in the absence of cultural experiments the conclusions 
can only be tentative: such evidence as is available, however, 
regarding their occurrence points strongly to the distinctness of the 
forms to be described. 
a V 
. i, rvcutulrews <j> 
J-.LoVaitcc 
Joem'irvoi 
&. , V uwi.t.'r strips 
Text-figs, a to d and f, show spetals of some of the forms of Silene maritima 
represented as flattened into one plane, e, diagrammatic longitudinal section 
of an hermaphrodite flower in the phase when the outer series of stamens (a 1 ) 
are ripe and the inner series (b 1 ) are still in the corolla tube. c\ coronal 
scale; P,lobelet. 
In all cases observed the different types of flower occurred on 
different plants, often widely separated and frequently found as 
isolated specimens amongst other forms. All but three of the flower 
