Let us again consider the case of two varieties each displaying the same character, but in the respective intensities _A_ and _a_. Each gamete of the _A_ variety bears _A_, and each gamete of the _a_ variety bears _a_. When they unite in fertilisation they form the zygote _Aa_. What will be its characters? The Mendelian teaching would reply that this can only be known by direct experiment with the two forms _A_ and _a_, and that the characters _A_ and _a_ perceived in those two forms or varieties need not give any indication as to the character of the zygote _Aa_. It may display the character _A_, or _a_, or a character half way between the two, or a character beyond _A_ or below _a_. The character of _Aa_ is not regarded as a _heritage_ transmitted to it by _A_ and by _a_, but as a character special and peculiar to _Aa_, just as NaCl is not a body half way between sodium and chlorine, or such that its properties can be predicted from or easily stated in terms of theirs. If a concrete case may help, a tall pea _A_ crossed with a dwarf _a_ often produces, not a plant having the height of either _A_ or _a_, but something _taller_ than the pure tall variety _A_. But if the case obeys the Mendelian principles--as does that here quoted--then it can be declared _first_ that the gametes of _Aa_ will not be bearers of the character proper to _Aa_; but, generally speaking, each gamete will either bear the pure _A_ character or the pure _a_ character. There will in fact be a redistribution of the characters brought in by the gametes which united to form the zygote _Aa_, such that each gamete of _Aa_ is pure, as the parental gametes were. _Secondly_ this redistribution will occur in such a way that, of the gametes produced by such _Aa_’s, on an average there will be equal numbers of _A_ gametes and of _a_ gametes. Consequently if _Aa_’s breed together, the new _A_ gametes may meet each other in fertilisation, forming a zygote _AA_, namely, the pure _A_ variety again; similarly two _a_ gametes may meet and form _aa_, or the pure _a_ variety again. But if an _A_ gamete meets an _a_ it will once more form _Aa_, with its special character. This _Aa_ is the hybrid, or “mule” form, or as I have elsewhere called it, the _heterozygote_, as distinguished from _AA_ or _aa_ the _homozygotes_. Similarly if the two gametes of two varieties distinguished by characters, _A_ and _B_, which cannot be described in terms of any common scale (such as for example the “rose” and “single” combs of fowls) unite in fertilisation, again the character of the mule form cannot be predicted. Before the experiment is made the “mule” may present _any_ form. Its character or properties can as yet be no more predicted than could those of the compounds of unknown elements before the discovery of the periodic law.