In dimorphic plants, such as Primula and Limim, two forms exist in about equal number and usually growing together. In one form the plants have flowers with a long style extending considerably beyond the short stamens, while in the other form, the position is reversed, the style being short and the stamens long. In the long-styled form the stigma is rough and fur nished with long papilla, and the pollen grains are small, while in the short-styled form the pa pilla of the stigma are short and the pollen grains are larger (Fig. 3). In the trimorphic hetero styled plants of Lythrum salicaria three lengths of styles are formed, long-styled, mid-styled and short-styled (Fig. 1). In each form the stamens exist in groups of two lengths corre sponding to the two other lengths in which the styles occur in other plants. The longest sta mens produce the largest pollen grains, the short est stamens produce the smallest grains. When insects visit the dimorphic and trimorphic flowers, their organs become dusted with the pollen at certain heights. When they later visit other plants with other lengths of styles, this pollen will be at the exact position and height to best cause cross-fertilization. In such cross ing a pistil always receives pollen from stamens of corresponding heights and the size of the pollen grain is thus proportional to the length of the style which its tube must traverse. Such crossing Darwin called legitimate. When a pis til of a dimorphic or trimorplic flower is crossed with pollen from stamens of different heights he termed it illegitimate fertilization. By very careful experiments Darwin found that only seeds produced as a result of legitimate crossing. give completely normal and fertile plants. Illegitimate crossing leads to the produc tion of progeny with all degrees of diminished sterility or even complete barrenness and give offspring which have all the characters of hy brids produced by the union of different species, Aside from the classes above mentioned a few plants are specially adapted to cross-ferti lization by small birds and snails, but such plants are few in number and their modifications are similar to those adopted by plants which are fer tilized by insects.
Prepotent Pollen.— The great of plants that have devices to secure cross-polltna tion also have some modification that ensures self-fertilization. This in a way would seem to have been developed as a safeguard to ensure seed development should cross-pollination fail to take place. In most cases the self-pollination takes place before or about the same time as the cross-pollination, and it would seem that in such cases where the plants are not self-sterile that a large majority of self-sterilized seeds would be formed. However, it has been found in many cases that the pollen of a different plant of the same race or species or in some cases of a different race will be prepotent over the plant's own pollen. In one instance, Darwin selected two flowers which had only recently opened on a plant of a variety of cabbage known as °Ragged Jack," and abundantly polli nated them with pollen from the same plant. After an interval of two or three hours pollen of a different variety, known as gEarly Barnes,* was dusted on the stigmas of the same flowers. Under the circumstances it would seem that little effect could be expected from the pollen of the Barnes cabbage, yet three out of the 15 plants raised from the seed formed by the above two flowers showed plainly that they were hybrids.
A similar experiment was carried out by the writer with cotton. A bud of Sea-Island cotton (Gossypium barbadense) was covered with a manila paper bag before it had opened. Early in the morning, when the flowers of cotton nor mally open and are pollinated, the bag was re moved and the stigma abundantly dusted with pollen from the same flower, after which the bag was replaced. Cotton is abundantly self fertilized, about 5 to 15 per cent of the flowers being normally cross-fertilized under the most favorable circumstances, so that this capsule should have set the normal number of seeds without further pollination. After four hours the bag was removed and the same sti dusted with pollen of upland cotton (G. Cr! baceutn), which belongs to a different but nearly related species. The seed of this Sea-Island cap sule gave five plants, of which three were clearly hybrids.
The prepotency of pollen can be easily ob served where different races or varieties are concerned, but in cases where the pollen of a different plant of the same race or species is prepotent over the plant's own pollen, as rs not infrequently the case, the fact is not so easy to prove. Darwin demonstrated prepotency in a number of cases of this lcind, using as his guide the superiority of seedlings raised from cross fertilized seed to those resulting from self-fer tilization, which after a few experiments can be used as a fairly safe index.
win's experiments with morning-glory (/Po Inca purpurea. The experiments were carried to the tenth generation and in each generation the height of the cross-fertilized plants greatly exceeded that of the self-fertilized plants (Fig 2). The ratios between the average he' ts of cross- and self-fertilized plants in the di erent generations were as follows: Benefits of Cross-fertilization.— The bene fit derived from cross-fertilization in the case of plants was first clearly brought forward through the classical treatise of Charles Dar win on "the Effects of Cross- and Self-ferti lization in the Vegetable Kingdom.* Sprengel at times apparently foresaw this law, but he seems never fo have grasped its full signifcance. In one place he states tit appears that nature has not willed that one flower should be fertilized by its own pollen,* yet he failed to realize that this carrying of pollen from one flower to an other was of any service to the plants them selves. Knight, Koelreuter and Herbert plainly had the main features of this law in mind, but did not recognize it as of sufficient importance to give it special attention. Darwin carried on extensive experiments to demonstrate the effect of cross- and self-fertilization in various plants and his conclusions are generally accepted to-day. His general plan of experimenting was to grow cross- and self-fertilized seed of the same plant in the same pot on opposite sides, with a parti tion between them. They were carefully watched and as often as one on each side germinated at the same time they were transplanted to an other pot and again placed on opposite sides of a superficial partition.