Nov., 1923] 
EAST — SELF- AND CROSS-STERILITY 
471 
tions will automatically reduce the number of segregating factors and hence 
the number of intra-sterile, inter-fertile classes. By taking advantage of 
pseudo-fertility it ought to be possible theoretically to obtain a population 
of plants belonging to a single class, a population in which every individual 
is cross-sterile with every other individual, though in this population each 
individual may fertilize the gametes or be fertilized by the gametes of other 
populations. These results, predictable on theoretical grounds, were 
actually obtained. A dozen or so populations with only two intra-sterile 
classes have been raised and studied, and two populations consisting of a 
single intra-sterile class have been found. Dr. Anderson has nearly com¬ 
pleted an analysis of the precise method by which the controlling factors 
are inherited. It does not seem advisable for me to discuss these results 
in detail here as they will be published shortly, but the main facts bearing 
on the general genetic problem of self-sterility can be stated in a very few 
words. 
First, the behavior of reciprocal crosses is the same. If the pollen of A 
is sterile on B , then the pollen of B is sterile on A. This is an important 
fact. In many of the populations studied, segregation and recombination 
of factors for pistil length was such that extremes of 25 mm. and 70 mm. 
were found. But under the usual conditions for carrying on the work, 
conditions under which pseudo-fertility was not a disturbing feature, 
incompatibility of the short-pistiled plant for pollen from the long-pistiled 
plant was just as marked as when the reverse cross was made. And, 
similarly, interclass crosses were just as easy to obtain when the long- 
pistiled plant was the mother as when the short-pistiled plant was used. 
Dr. Anderson believes that this result is aided by a high positive correlation 
between the length of “life” of the flower and the length of its pistil. Be 
that as it may, the growth rate of the pollen tubes in incompatible matings 
is so slow that even if the flower from a short-pistiled plant remained on the 
stem for as long a period as with its long-pistiled sister, the tube does not 
have the opportunity to traverse more than two-thirds of the pistil distance. 
It is possible, however, so to control conditions that one may sometimes 
obtain seed from a particular mating when it is impossible to obtain it after 
a mating made the reverse way. For example, if plant A , a plant in vigorous 
condition and at the height of its flowering season, is used as the pollen 
parent on an incompatible plant B , a plant in weakened condition at the 
extreme end of its flowering season, some seed will be obtained as a mani¬ 
festation of pseudo-fertility; but the reciprocal cross is impossible. 
A further conclusion may be derived in part from the fact that reciprocal 
matings are identical when not interfered with by external conditions whose 
effects are fairly well known and for which reason are controllable and their 
results predictable. It is to the effect that, though the gametes formed by 
a particular plant may be packed with many different combinations of 
genes, as proven by the variable population of zygotes produced, as far as 
