cate flowers. Small flowers, as in sugar beets, wheat, alfalfa, 
mignonette, are hard to see enough of to work with, while 
delicate flowers, such as are characteristic of touch-me-nots, 
melons and many greenhouse plants, will not tolerate much 
mutilation and are prone to fall off unless handled with ex- 
treme care. Some plants, as corn for example, are so bountiful 
with their pollen dust that it is extremely hard to keep the 
female flowers or “silks” protected from it, even by the most 
careful manipulation of paper bags and sterilizing apparatus. 
Carnations and many double flowering plants, on the other 
hand, have so little pollen that extreme difficulty is experienced 
in securing enough with which to work. Then again, there are 
plants, such as many varieties of fruit trees, that will not 
produce seed if pollen is used on the pistils of flowers of that 
same plant. In some cases this is so marked as to lead us to say 
that pollen from the same plant used on the pistils of that plant 
actually poisons the whole floral structure. On the other hand, 
garden peas, sweet peas, and the inconspicuous green flowers of 
the violet normally pollinate themselves, i. e., the little pollen 
bags of the flowers of these plants burst open and shed their dust 
on the already sticky pistil before the flower has passed the bud 
stage. When the flower opens, a sort of an April-fool joke is played 
on the rest of Nature, because its little seeds are already on their 
way to becoming grown up. The ideal material for the study of 
our laws of heredity, so far as our present knowledge goes, is a 
plant with flowers normally self-pollinating, of fair size — flowers 
that will stand mutilating and bagging without undue tendency to 
fall, flowers which produce a large quantity of seed from a single 
flower, and flowers with a long blooming period. Plants easy to 
raise, which give' two or three generations a year, are additional 
recommendations, as every one is impatient for results. There are 
many such plants, but perhaps the garden pea is one of the com- 
monest and has most of these desirable attributes. The lily lacks 
a great many of these virtues, and one of the worst drawbacks to 
its use is the length of time (two to five years) that is required to 
secure a blooming plant from seed. By the time a generation 
grows up, one has lost interest in it. Because the garden pea was 
such ideal material, Mendel, the father of the science of genetics, 
used it for the working out of his problems. Had he used such 
plants as dandelions, evening-primroses and corn, I fear we 
would still be compelled to admit that our knowledge of heredity 
is a hopeless chaos, and heredity itself — “a maze in which science 
loses itself.” 
With such a plant as the pea, we start our crossing work, 
and the tools we need are mostly innate attributes — infinite 
care and patience, and, in addition, a few paper bags or cel- 
