326 REV. GEORGE HENSLOW ON THE 
them ; still the abundance of pollen renders the flowers very liable to be self-fertilized. 
The innermost stamens develop first, while the outer ones are delayed. This is contrary 
to the order in Buttercups. Perhaps it is correlated to the elongated styles, which 
project outwards, and whose stigmas will thus be fertilized more probably by the last or 
outermost dehiscing anthers, just as the carpels are in Buttereups, though in this latter 
case by the innermost stamens. In addition, therefore, to the probability of self- 
fertilization occurring in the expanded flower, the withering of the corolla which binds 
stigmas and anthers together will inevitably ensure it. 
Many other flowers retain their corollas or perianths for some time in a withered or 
* marcescent ' condition, such as Gentianee, Campanula, Ze, and I would suggest the 
probability of such being in all cases at least an aid to secure self-fertilization. 
4. ii. The excision of the corolla, and stamens in part(?). It occurred to me on 
observing, I may say, the almost universal tendency to reduce the size of the corolla 
or to suppress it altogether when flowers are habitually or solely self-fertilized, and 
that it was also the case with the female flowers of polygamous species as well as of 
* gynodiecious" plants, as Mr. Darwin calls them (* Forms of Flowers,’ pp. 287-309), 
that there might be some sort of compensating process between the pistil and the 
corolla—that when the latter is arrested the former grows rapidly, so that if it were 
early cut away a normally proterandrous flower might become homogamous ; but experi- 
ments with strongly proterandrous flowers gave mostly negative results when the corolla 
alone was removed from the bud. Hence I am led to regard the reduction of the corolla 
as a secondary result issuing from some tendency to check the energy of the stamens, 
which may convert a proterandrous flower into a self-fertilizing one, or a normally 
hermaphrodite into a female. The reduced corolla, and sometimes a reduced calyx *, 
must therefore be regarded as correlative results issuing from certain hidden causes 
affecting the essential organs.. I have tried several experiments of removing the corolla 
both with and without some or all of the stamens in Pelargonium; the style and 
stigma then appeared to gain vigour, and to mature from two to five days earlier than 
usual. 'lhis, however, requires corroboration. 
Mr. Darwin records the fact that certain proterandrous or self-sterile flowers, which 
had their corollas removed, set seed, such. as Delphinium Consolida and Viola tricolor 
(* Cross and Self-fertilisation,’ p. 420, note) ; and this led me to regard ‘such cases as E 
strengthening my deduction; but I am now inclined to withdraw it, and substitute the _ 
belief that the reduction of the corolla is an outward index, as it were, of some con- — 
stitutional peculiarities which are affecting more especially the stamens. I have else- A 
where observed of diplostemonous flowers that if a whorl of stamens be suppressed itis 4 
usually that opposite the petals, and in the Paronychiew the corolla often goes with it, 
showing some hidden bond of connexion between these two whorls. Hence I would 
-agree with Mr. Darwin, who thus speaks of the smallness of the corolla of the female ; 
forms of usually hermaphrodite plants :—‘ It seems therefore probable that the de. ` 
* Rhamnus catharticus (t Forms of Flowers,’ p. 308) and Amphicarpac monoica have cleistogamous flowers without 
even “ the semblance of a e e (I. a) 
