The seeds from such flowers would be strong and would have the best chance to survive. The plants that grew from these seeds would also inherit the tendency to secrete sweet juices near the flower.

In probing for sweets, the insect would irritate the parts it touched, and this would cause an extra flow of sap there and very likely the manufacture of more sweet juice; so the nectary came to be developed.

You can understand how this might be by recalling how the skin of your hand changes when you first try to do some new and hard work, like rowing a boat.

After you have rowed a little while your hand is blistered. The constant rubbing of the oar in one place has irritated it, just as you can imagine the tongues of the insects rubbing against the delicate flower tissue would irritate it. Wherever a place on the skin is irritated, the blood flows to that spot; and so in the plant, where it is irritated, there will likely be a collection of sap. After the blood has flowed to the place on your hand which was rubbed by the oar, the spot becomes red and inflamed and pains you, and finally the skin separates in the form of a blister and a new skin forms underneath; and if you keep on rowing, your hand does not keep on blistering, but actually makes a new kind of skin to protect the rubbed places, and what we call a “callous” or hard spot is formed. The skin is many times thicker here than elsewhere, and was formed on purpose to protect the place. So we can understand how irritation might change a plant organ and in time form a nectary.

But how about petals, you are asking. Well, imagine yourself in those old times when plants made their first flowers out of pistils and stamens only.

These primitive flowers were probably not very showy. Primitive flowers means first flowers,—flowers that lived way back in the beginning of plant life.

They had no petals, but they secreted juices which the insects liked. Those early insects were queer fellows, too, not very much like our insects, except that they were fond of sweets and liked to eat the tender parts of the flowers, just as our insects do to-day. They ate nectar when they could find it and did not disdain pollen, which, it is to be feared, they sometimes ate, anther and all; and, what is worse, they in all probability frequently dined on pistil, which was very bad for the plant.

Now imagine one strong plant secreting a good deal of nectar. The insects would be likely to eat this and let the pollen and pistil alone, only in getting to the nectar, they would be apt to dust the pistil with pollen from another plant which they had been visiting and would also brush off some pollen against their bodies.

Thus the strong plant with the abundant nectar would be cross-fertilized and would keep its pistil unharmed. It would be very likely to develop good strong seeds that would grow and again bear strong flowers with plenty of nectar. Now, remember the essential organs—that is, stamens and pistil—seem to find it a little easier to change than other parts of the plant; so it would not be surprising if in time some of the stamens were to become different. You see, the insects in visiting the flowers would irritate them more or less walking over them and clinging to them, and they would be likely to undergo change for this reason; and if it happened that in some flower a row of stamens got too full of sap to know what to do with themselves and so spread out a little broader and more leaf-like and kept their yellow stamen color or bleached-out white, that flower would be seen far and near and the insects would go straight to it, for insects have the sharpest kind of eyes for seeing bright colors a long way off. You see what would happen; all the flowers whose stamens had done so would be abundantly cross-fertilized,—that is, all their seeds would get fresh pollen from another strong plant, and the plants growing from these seeds would inherit the tendency of their parents to form petal-like parts from some of the stamens. The flower could well afford to lose part of its stamens for this purpose. Of course as time went on, these stamens, which were half petals, might develop more and more in the direction of signals,—that is, might become more and more perfect petals, finally losing all trace of their old life as stamens.

Of course no one can say that is just the way it came about, but it is likely that in some such way it happened, for there are proofs of it which you may like to read when you grow older.