BREEDING STAPLE FOOD PLANTS. 261 
produced to secure one superior one. A prominent Potato breeder has 
said that not more than one new variety in two thousand proved 
sufficiently valuable to become a commercial variety; yet the breeding 
of Potatos has added sufficient to the value of that crop to pay for the 
required labour a thousand, if not a million, times over. 
The above-mentioned experiments in Wheat breeding are reported 
upon in full in Bulletin No. 62 of the Minnesota Experiment Station, 
and a number of copies are held in reserve for exchange for writings on 
plant and animal breeding. 
STEPs IN VARIETAL KH VOLUTION. 
Many amateur hybridisers make a mistake in thinking that the 
hybrid is the end sought, the difficult feat to be performed, the result 
which brings value and creditable mention. Only when that number 
of hybrids is produced that a really useful or attractive form is secured 
and given to the world are the results valuable from an economic or 
artistic standpoint ; and only when the method is so recorded as to be a 
suide for future work are the results valuable from a scientific stand- 
point. In the case of many crosses, the individual plants, of hybrids between 
dissimilar species, mostly revert downward in yield, or in other intrinsic 
qualities. It is only the exceptional plant which varies strongly in the de- 
sired directions. Extensive selection must therefore accompany crossing 
to get the highest possible results. The selection should begin with the 
varieties to be crossed. How little we know as to the average results of 
the crosses between any two species or varieties, and how little are we 
able to predict how many individual plants of each cross we must throw 
away for each useful one we shall find under rigid selection! There 
would seem to be no end to this portion of botanical research. 
Selection of the parent plant within the variety is ofttimes very 
important. In Wheat, for example, there is great variation in the 
strength and in the prepotency of the individual plant, and it is im- 
portant that good plants be chosen and that the power of these be tried 
to see that they are potent in the production of good plants; in other 
words we must raise a generation of plants from each parent plant, and 
compare these second generations or broods that we may know the 
relative values of each of the mother or father plants as parental stocks. 
We must measure the breeding value of the parent plants by judging or 
comparing their produce. The breeder of pigs likes to have the male, 
which is to head his herd, show his pedigree on his back ; he also requires 
a good written pedigree, but the value of the sire or dam is still better 
judged by seeing a large number of his or her progeny. 
Since the labour of developing hybrids, when fully carried out, often is 
very great, the probabilities of useful results from crosses should, where 
practicable, be increased by carefully selecting the foundation or parental 
individuals. Since variation may occur all along the line, where it is 
practicable, the best flower should be chosen on the best part of the best 
plant. 
Hybridising, in breaking up the centripetal force of heredity and by 
letting loose the centrifugal forces of past generations—which forces, acting 
