(1) The purity of the gametes in regard to certain characters.
(2) The distinction of all zygotes according as they are or are not
formed by the union of like or unlike gametes. In the former case,
apart from Variation, they breed true when mated with their like; in
the latter case their offspring, collectively, will be heterogeneous.
(3) If the zygote be formed by the union of dissimilar gametes, we
may meet the phenomenon of (_a_) dominant and recessive characters;
(_b_) a blend form; (_c_) a form distinct from either parent, often
reversionary[20].
[20] This fact sufficiently indicates the difficulties involved in a
superficial treatment of the phenomenon of reversion. To call such
reversions as those named above “returns to ancestral type” would be,
if more than a descriptive phrase were intended, quite misleading. It
is not the ancestral _type_ that has come back, but something else
has come in its guise, as the offspring presently prove. For the
first time we thus begin to get a rationale of “reversion.”
But there are additional and even more significant deductions from the
facts. We have seen that the gametes are differentiated in respect
of pure characters. Of these pure characters there may _conceivably_
be any number associated together in one organism. In the pea Mendel
detected at least seven--not all seen by him combined in the same
plant, but there is every likelihood that they are all capable of being
thus combined.
Each such character, which is capable of being dissociated or replaced
by its contrary, must henceforth be conceived of as a distinct
_unit-character_; and as we know that the several unit-characters are
of such a nature that any one of them is capable of independently
displacing or being displaced by one or more alternative characters
taken singly, we may recognize this fact by naming such unit-characters
_allelomorphs_. So far, we know very little of any allelomorphs
existing otherwise than as _pairs_ of contraries, but this is probably
merely due to experimental limitations and the rudimentary state of our
knowledge.
In one case (combs of fowls) we know three characters, _pea_ comb,
_rose_ comb and _single_ comb; of which _pea_ and _single_, or _rose_
and _single_, behave towards each other as a pair of allelomorphs, but
of the behaviour of _pea_ and _rose_ towards each other we know as yet
nothing.
We have no reason as yet for affirming that any phenomenon properly
described as _displacement_ of one allelomorph by another occurs,
though the metaphor may be a useful one. In all cases where _dominance_
has been perceived, we can affirm that the members of the allelomorphic
pair stand to each other in a relation the nature of which we are as
yet wholly unable to apprehend or illustrate.