188 THE REV. С. HENSLOW ON THE ORIGIN OF FLORAL ASTIVATIONS. 
whorls by a continuous spiral growth, so also, in the order of development of the whorls 
themselves, it is by.no means always true that they appear in succession acropetally. 
Thus, as a consequence of a fiower always terminating an axis, it is the fact that not 
infrequently there is subsequent intercalation of whorls exterior to some already formed. 
Thus in Silenec “the first whorl of stamens which appears is opposite the sepals, while 
the second is produced opposite the petals, and (according to M. Chatin, but my own 
observations scarcely corroborate the statement) exterior to the first-produced whorls ”%, 
As far, however, as Stellaria holostea is concerned, the stamens opposite the petals seem 
to me to be intercalated between the first-formed stamens; so that the andrcecium forms 
one circle only. The petals, however, are decidedly later than both sets of stamens, the 
latter having attained considerable dimensions when the former consists of minute semi- 
circular plates. This subsequent intercalation of whorls is perhaps due to what one may 
call “developmental energy” being stopped in its axial course, and, thus diverted, 
evineing itself by lateral or lower outgrowths. "This also helps to explain the acrofugal 
order of expansion in Adoxa (see Note В, p. 194). 
2. Irregular flowers.—There is, at least, a presumable probability, based upon what is 
known regarding the mutual adaptation of flowers to insects, that all irregular flowers 
have originally arisen by such adaptations, and that the varieties called “ peloria” are 
reversions to the regularity of the primitive type Т. Now, in order to become irregular, 
one obvious process is the greater development of one or more parts of a whorl, often 
at the expense of the others, which remain more or less dwarfed; so that it would be 
presumable 4 priori that irregularity would determine which part or parts should arise 
first from the receptacle. But this does not always appear to be the case, though at 
present I am not aware that any law has been discovered which regulates the order of 
growth. The fact that any adult part is larger than another part does not necessarily 
imply that it started first. Thus is it often with those anthers which appear first, and 
far outstrip the petals in size, as in Stellaria; yet the petals, when adult, often exceed 
the stamens immensely, as in the last-named genus and many others. A like remark 
may be made with reference to stipules of trees, as of the beech, which in the bud grow 
to four times the size of the leaf to which they belong, yet soon cease to grow, when 
the leaf ultimately exceeds them many hundred times. 
М. Chatin (/.¢.) also observes that the polystemonous andreecia of Mesembryanthe- 
mum, Opuntia, Cactus, &c. are developed centrifugally ; but the inner, older members 
become partially arrested, and the outer are the first to dehisce their anthers. 
* “Some Alsineæ,” that author observes, “ with reduced flowers, give rise only to the whorl opposite the sepals. 
This is the rule in Paronychiacew, where sometimes the abortion of the stamens opposite to the petals is accompanied 
by the abortion of the petals themselves ? (Bull. de la Soc. Bot. de France, C, 1874, р. 133). А propos of this fact, I 
would observe that these and other reduced flowers are mostly, if not all, self-fertilizing ; and besides the degrada- 
tion of the petals, as being no longer required for attracting insects, the number of stamens gets reduced, ог 
more or less aborted; thus is it in Senebiera didyma, Stellaria media, Viola canina, and other species with 
cleistogamous flowers ; sometimes the andrccium is even barren, as in some early-developed individuals of Capsella 
Bursa-pastoris, Glechoma hederacea, &c. 
t The order of “ emergence” of the whorls of Linaria vulgaris is as follows—calyx, corolla (Ұма is quite 
—M— —— lastly, the pistil. 
