164 
TOPULAR SCIENCE REVIEW. 
and that the production of the ray was merely a secondary 
result of this, the spare material necessarily appearing in some 
shape or other. But this view is irreconcilable with the fact, 
that in some capitals (e.g. Centaurea Cyanus) the suppression 
includes not only stamens but pistils. Here there is a large 
unhandsome ray, and yet no monoecious condition. What possible 
interpretation can be given of such a ray of neutral flowers, 
excepting that it is intended to attract insects ? 
We have seen that in such flowers as Feverfew self-fertilisa- 
tion must not be an infrequent occurrence, though less frequent 
than intercrossing. In some other nearly allied flowers — for 
instance, in the Marigold — it is altogether prevented. For 
here, not only do the ray-flowers cease, as in Feverfew, to pro- 
duce pollen, but the disk-flowers develop no stigmas. It is 
eurious to notice what modifications this further step entails. 
The pistils of the ray, as before, are without unnecessary tufts 
of hair and bear their stigmas on a bifid extremity. The disk- 
flowers have their anthers as usual united into a tube : now, 
were the pistil altogether absent, there would be no means of 
bringing the pollen to the surface. The style, therefore, remains 
with its terminal brush ; but as there are no stigmas this tufted 
extremity no longer requires to be divided. It is therefore 
simple and not bifid. 
The two capitals I have described are from one group of 
composites ( Senecioicls ). In the other groups the main facts 
are the same : that is to say, the anthers in all form a tube 
from which the growing style brushes out the pollen ; in all 
the pollen is spread on the surface before the stigma of the 
same flower is mature, and in all there are collected in one 
head flowers in different stages, so that in all intercrossing by 
aid of insects must be the rule. But the position and arrange- 
ment of the hairs on the style which sweep out the pollen differ 
>n each group ; and it is this difference which furnishes the 
characteristic mark of each subdivision. 
I pass on to one of the most attractive of irregular forms — the 
Papilionaceous. The whole purport of this is to ensure the 
transfer of the pollen to an insect, which is the same thing 
as to ensure repeated intercrossing. I will take as an example 
the common furze ( Ulex ). The upper petal — standard — re- 
quires little notice. It serves, by its size and brilliancy, to add 
to the attractiveness of the flower, and also as a protection to 
the more important parts below. The two lower petals are 
united together to form a kind of boat — carina — in which are 
lodged the pistil and stamens. The comparison to a boat is 
rather far-fetched, for not only is the casina closed below in the 
part which would represent the keel, but also above ; so that 
the organs within are close prisoners. Were the upper edge of 
