34 ILLINOIS ACADEMY OF SCIENCE. 
ing touches upon forms already well-developed. It does not pro- 
duce new forms, but combines old ones. The combination may be 
very desirable, but the range is restricted, because the process is 
always working over the old material. 
There is a still more fundamental limitation to the usefulness 
of hybridization as a permanently useful process. Mendel's law 
has shown that the progeny of hybrids split up in certain simple 
and definite ratios. In many cases, one half of these plants of 
the second generation are pure parent forms, and only one half 
retain the hybrid combination, and so on for each succeeding 
generation. This means that approximately only one half of the 
progeny of hybrids come true, a percentage of loss that does not 
commend the process as one of permanent value. However, there 
are two facts that save the situation and may save the process as 
a useful adjunct to more important ones. A few hybrids have 
been developed that do not seem to split; this is notably true of 
Burbank's famous triple hybrid, the Shasta daisy. If it is true 
that some hybrids are not Mendelian, such can be made perma- 
nently useful. 
The other saving fact is much more important. Very many 
cultivated plants are not propagated by seeds, and therefore their 
hybrids are not subject to the Mendelian splitting. Propagation 
by bulbs, tubers, slips, grafting, etc., includes such a wide range 
of plants that the sphere of useful hybridization is extensive 
enough. It will be noted that most of Burbank's combinations 
are plants that fall within this category. It touches flowers and 
fruits, chiefly; that is floriculture and horticulture; but not the 
far more important field of agriculture, where lie the fundamental 
problems of practical plant-breeding. 
Within the last three or four years, a remarkable field of 
hybridization has been opened by the researches of Winkler and 
Baur. Ordinary hybridization is effected by the sexual method, 
which means actual fusion of parental cells, and the development 
of the hybrid from a single fusion cell. All of the behavior of 
hybrids noted above is based upon this method. 
The new field introduces the production of hybrids by grafting. 
The influence of the stock on the scion is as old a question as is 
the process of grafting, but the recent researches referred to 
have uncovered the situation. If a hybrid is to be defined as the 
result of the fusion of cells from dissimilar parents, then it 
remains to be proved whether graft hybrids are true hybrids. 
But if it is to be defined as a form exhibiting a combination of 
