THE REV. G. HENSLOW ON THE ORIGIN OF FLORAL JESTIVATIONS. 193 
numbered quincuncially, No. 1 of each whorl being nearest that of the preceding, and 
on the same side of it, and then if we strike out all those, say, numbered 5, the re- 
maining four parts in each whorl, on closing up the gaps, instantly form regularly 
alternating whorls of fours. Thus, then, I conceive the cruciferous flower to have 
arisen by а symmetrical reduction from fives to fours—first by the arrest of develop- 
ment of every fifth part of quincuncially arranged whorls, then by an additional arrest of 
a posterior and anterior stamen of the outer whorl, as well as a partial arrest ofthe 
carpels as described above. 
A botanist who accepts the idea of a Crucifer being referable to a binary symmetry, is 
M. Victor Meschajeff*. He regards the two stamens constituting one of the taller pairs 
(as also of two adjacent petals) as reversions to a primitive type, and considers that the 
usual free condition of the stamens is a result of dédoublement. 
I would take this opportunity of expressing my belief that this is a fundamentally 
erroneous view. It seems, too, to agree with Dr. Masters's expression of “ inseparate" 
as a substitute for “cohesion.” If cohesion, however, be objected to as implying а 
previous condition of freedom which never existed in petals normally joined, so “ inse- 
parate" would seem to imply that union was prior to freedom. There could be no 
reasonable objection, as far as I see, to the term “ inseparate," provided no explanation 
of the condition is implied, but simply the fact that the parts are “ not separate." Dr. 
Masters, however, qualifies the term by saying that the parts thus described as inse- 
parate are not separated. The use of this participle at once introduces а new idea, 
and implies the existence of some tendency ѓо separate what was before joined. Now 
this seems quite as objectionable as the term cohesion seems to that author (* Botany 
for Beginners,’ p. 27) f. 
But, according to the laws of evolution, integration, where it exists, is always sub- 
sequent to isolation or freedom. The rare cases of * dialysis," as of Convolvulus and 
Campanula with free petals, or of ** solution," as of Pyrus and umbelliferous flowers with 
an inferior calyx, would be cases of reversion ; while, in the Crucifere, freedom of parts, 
except іп the gynecium, is clearly a fundamental condition, as it has no affinity with any 
gamopetalous order: consequently the occasional cohesion of petals or stamens in any 
cruciferous flower is not a reversion but a more advanced differentiated state, and 
appears to be due simply to the close approximation of the parts thus occasionally 
joined. ! 
If such a flower should ever have an apocarpous pistil, thai would be a true instance 
of reversion ; but no case of union could possibly be such. Vocal 
For the above reasons I have always looked doubtfully on “ chorisis ” аза principle i 
flowers. Doubling or an increase of number of parts is common enough, as indeed 
* Bull. de la Soc. Imp. des Nat. de Moscou, 1872, no. 2, quoted in Bull. de la Soc. Bot. de France, A, nis p. 16. 
T Dr. Masters has been gcod enough to give me the following comment on these remarks el * Inseparate' I apply 
to cases where separation of parts, though * potential, has not yet occurred, — — if the course of develop- 
ment be arrested, it may never actually occur.” But it is this very ‘potentiality, ы tendency to "t parts, 
Which I can only recognize as identical with atavism or reversion. Hence I think it ought not to be considered a 
principle of development, but rather а case of retrogression, and therefore exceptional. 
SECOND SERIES.—BOTANY, VOL. I. 2E 
