In practical animal breeding the blended inheritance just described is not very useful, for even though a blended character might appear which is just what the breeder has been looking for, it will not occur in more than half the offspring and can not ever be depended on to show itself in any particular individual. This explains largely why pure-bred stock is always more desirable than hybrid, and why breeders strive so eagerly to obtain desired traits in pure-bred animals. In plants, blended characters are much more valuable, for two reasons; first, because the offspring are so numerous that even though half of them come out pure, and so lack the desired blend, there are enough left that have it to make the crop worth while; and second, because propagation by cuttings is possible in very many kinds of plants, which means that the same plant is kept going in hundreds of places, and for tens or even hundreds of years. A trait that is desirable can be perpetuated indefinitely by this means, even though it may be a blending of several hereditary traits, which would separate out in a few generations by ordinary means of propagation. There are several more things in heredity that must be taken up while we are on the subject, so we shall have to return to the chromosomes for a while. We have seen that there are several determiners to each chromosome; for convenience, we assigned three apiece to our chromosomes, except the ninth, which has to get along with two; but in reality the number to each chromosome is often much greater. This grouping of the determiners, several to a chromosome, carries an interesting consequence with it, in that all the hereditary traits controlled by one chromosome have to go together in reproduction. In the example we have already used A, B, and C are together; therefore any individual that shows character A must show B and C as well. The most striking instances of this are certain traits that are bound up with sex, but we cannot describe these further until we have looked into the heredity of sex, which we shall do in a minute. First a word must be said about occasional exceptions that turn up to the rule that we have just stated. In the study of thousands of specimens now and then one has been found in which there has evidently been a swapping about of determiners. We can illustrate the situation by supposing chromosome one is found to contain determiners A, b, and C, instead of A, B, and C; one small-letter determiner has traded places with a large. Of course, the effect of this is to permit different combinations of hereditary traits than ordinarily occur, and at the present time students of heredity are actively engaged in following this up to see how it happens, and what advantage can be taken of it. This crossing over of determiners from one chromosome to another takes place only among such as are actually in contact at times within the nucleus as seen under the microscope, which