POLLINATION. 206 POLLINATION. ierogyny; if the pollen grains are slied first, •tn-otnndry or pnitcrundry (I'lg. 4). In general, ero-i-pollination seems to be some- what 'iiilvantageous to plants becanse it is sup- posed to prevent close inbreeding, whieh m both plants and animals .seems to result m de- eeneracv. The reason for this is not known. There are varioiis devices by which cross-pol- lination is obtained. The most perfect device is the so-called diceciou.s habit illustrated by the willows (Fig. 2|. In the so-called niono'cious habit (Fig. l") • cross-pollinatiou is less certain. transferred by insects. The structure of the insect-pollinated Mowers is commonly more or less complicated and diflfers somewhat widely Fig. 2. dkecism. Catkins of the crack willow {Salix &agills). the lower Btaininate. the upper pistillat*. These catkins are boriw on separate plants, thus insuring cross-polllnation. from that of flowers pollinated in other ways (Fig. 5). For example, the coroUas are com- monly prominent and highly colore<l, and it is be- Fu;. 1. S(OS<EC18.M. The ha7.el (Corvlus). showinc pistillate flower clusters (6) and staiulnate catkins (a) on the same twig. Self-pollina- tion is possible here, as is never the case In a dioecious plant like the willow (Fig. 2). though it is often facilitated by the pistillate flowers lioing uppermost on the tree. In the case of insect pollinated flowers cross-pollination is sometimes secured by protandry ( Fig. 4 ) , or pro- togyny. In many cases the stigma when mature is higher than the stamens and hence self-pollina- tion is dinicult. Perhaps the most etTcctive de- vice to insure cross-pollination is the impo- tence of the pollen upon the stigma of the same flower and even of other flowers on the same plant. It must he remembered that in some plants self-pollination is not only possible, but even common. It is invariably the case in cleistogamous flowers. (See CLEisTOG.>tY.) In the subterranean flower of the violet self-pollina- tion is necessary, and the rich development of seeds shows that it is effective. Pollination is effected in various ways in plants, among which the most important are the follow- in": ll) Insect pollination. In a vast number, perhaps even the majority of plants, pollen is Fro. 3. DIMOBPHISM. Flowers of the Houstonia, illustrating a means for facil- itating rross-pollination. In a the styles are hnm aaa the stamen fllameuts short ; while in b the reverse is true. PoilPii nlitaiued from anthers on long stuuieiis is likely to hebni.-he.i oft upon the stigmas of the InnK st.vles ; while pollen from the short stamens will be brushed off upon the stigmatic surface of the short styles. lieved that insects are more likely to visit flowers if their corollas are colored. Some experiments, however, would seem to show that color may not serve conspicuously as an attraction. Many
