the forming of the petals comes about through a kind of expansion, the secondary corollas on the other hand are formed by contraction, that is to say, in just the same way as the stamens. And so we see within perfectly expanded corollas smaller, contracted, secondary ones, as for example in the Narcissus, in Nerium and in Agrostemma.
56
Furthermore, in various species we see alterations which are still more striking and remarkable. In some flowers we notice a little hollow filled with honey-like juice at the inner base of the petals. This little cavity becomes deeper in some species and types than in others and produces on the back of the petal a spur or horn-like protuberance, the shape of the rest of the petal being at the same time more or less modified. This can be distinctly seen in many types and variations of Aquilegia.
57
We find the nectary most transformed in Aconitum and in Nigella , but even here only a little attention will enable us to see its resemblance to a leaf. In Nigella especially it tends to grow into a leaf or petal and through the transformation of the nectaries the flower becomes double. In Aconitum it is easy to see the resemblance of the nectaries to the domed-shaped petals under which they are hidden.
58
Having already said that the nectaries are a transitional stage between petals and stamens we may at this point say a few words about irregular flowers. The five outer petals of Melianthus , for example, could be described as true petals, but the five inner ones as a secondary' corolla consisting of six nectaries, of which the uppermost nearly resembles the leaf-form while the lower—which is indeed called the nectary—differs most from it. In this same sense the carina or keel of the Papi-lionaceae might be called a nectary in so far as it, of all the petals of this flower, is nearest in form to the stamens and differs greatly from the leaf-like form of the so-called vexilli. In this way too we may easily explain the brush-like appendages which are attached to the end of the carina of some species of Polygala : so shall we form a distinct idea of the real meaning of these organs.
59
It is hardly necessary to avow that these remarks are not intended to bring into confusion all that has hitherto been separated and classified through the endeavours of observers and systematists. We only wish to explain more clearly the variable formations and developments of the plant-kingdom.
C HAPTER VIII
MORE ABOUT THE STAMENS
60
Microscopic observations have placed it beyond doubt that the generative organs of the plants as well as other organs are produced by spiral vessels. We take this as a basis for the argument that the different parts of the plant, which have so far manifested themselves to us in such varied forms, are none the less intrinsically the same.
61
Now as the spiral vessels are situated in the middle of the bundles (vascular bundles) and are enclosed by them, we can to some extent come to a better understanding of the above-mentioned strong contracting force if we imagine the spiral vessels, which really look like elastic springs, exerting their utmost power so that they overcome the expansive tendency of the sap-vessels.
62
The shortened vascular bundles can then no longer expand; they are not able to unite so as to form a network by anastomosis, and the cellular tissue, which otherwise fills up the spaces of the network, can no longer develop. Here, all the causes for the expansion of stem-leaves, calix and petals are at an end, and there appears a frail, extremely simple thread or filament.
63
Hardly are the fine little membranes of the anther formed, than the extremely delicate vessels terminate in them. Now if we admit that here the very same vessels, which otherwise become lengthened and expanded and united with one another, are at this stage in an extremely contracted condition; if, moreover, we see coming from them the highly developed pollen which through its active energy