Mutations and Evolution. 
71 
shortening the styles, slightly broadening the leaves and bracts, 
making the bud cones a little thicker and the sepal tips somewhat 
shorter. The resulting segregation in F 2 is sharp and complete, 
but the percentage of brevistylis is usually below expectation in all 
crosses, owing to brevistylis zygotes being apparently less viable. 
This is a very good example of a type which was able to 
survive for an indefinite period in competition with its parent form, 
because while producing practically no viable seeds it was 
perpetuated by occasional crosses of its pollen giving rise to 
heterozygous latnarckiana plants. 
Mutations Arising at or After Fertilization. 
De Vries assumes that all mutations take place in the formation 
of the germ cells, that is presumably at the time of the meiotic 
divisions, and there is a great deal of evidence that many 
germinal changes arise at this time. During this period the 
pairs of chromosomes are being redistributed, and the intricate 
manoeuvres of the chromatin material just previous to this 
separation of the chromosomes gives an opportunity for a variety 
of rearrangements and aberrant distributions of particular chromatin 
particles. That the hereditary behaviour known as crossing-overt 
by which part of the factors which were in the same chromosome 
cross over to its mate, has its basis in changes occurring at this time 
has been shown experimentally for Drosophila by Plough (1917). 
He found that by subjecting the females to high or low temperature 
the amount of crossing-over was increased. Further experiment 
showed that the critical period, during which crossing-over in the 
offspring was affected, was during the period of conjugation of the 
chromosomes in the maturation of the eggs. From this it follows 
that crossing-over takes place at that time and not earlier. 
In the GEnotheras it has been shown that many of the 
irregularities leading to new chromosome numbers arise during 
meiosis. The origin of mutations with an extra chromosome and 
the phenomena of non-disjunction in Drosophilaj of the sex- 
chromosomes (failure to separate in meiosis), are cases in point ; 
also the origin of the supernumerary chromosomes in Metapodius 
(Wilson 1909), which arise through irregularities in chromosome 
distribution during spermatogenesis. In addition, there is evidence 
both in plants and animals that germ cells with a double number of 
chromosomes are occasionally produced, through an alteration in 
the meiotic phenomena. These matters will be discussed in 
connection with triploidy and tetraploidy. It will be found that the 
