Mutations and Evolution. 
87 
The determinations of chromosome numbers from root tips by Miss 
Lutz are therefore probably correct, although it is certainly 
desirable that they should be confirmed by study of the germ cells. 
Although Hance regards the diminutive extra chromosomes of 
Miss Lutz in (E. rubrinervis De Vries and (E . mut. aberrans as 
mere temporary fragments, yet the evidence seems to indicate 
that they are permanent products of a fragmented meiotic 
chromosome dividing by mitosis, and therefore present in every 
nucleus. Their constant difference in size in the two forms also 
supports this view. 
Summing up the results regarding 15-chromosome forms of 
(Enothera , we may conclude that while several of the types have 
probably arisen from lamarckiana through having different 
chromosomes as the extra one, yet several others, which have 
appeared only in the offspring of lata x lamarckiana or of lata 
selfed, evidently have a closer relationship to lata , than to any 
other form, and semilata Gates at least probably has the same 
extra chromosome. The differences between the other forms may 
be accounted for partly by the presence of other factors or mutant 
characters in addition to the extra chromosome, and partly by 
fresh rearrangements of the 15 chromosomes. 
The same irregular meiotic division, or non-disjunction of a 
pair of chromosomes, also takes place no doubt in other forms, and 
may be expected to lead to the appearance of occasional aberrant 
members in other species. Winge (1917) apparently found such 
a case in Staphylea pinnata , which has 12 pairs of chromosomes, but 
one haploid group with 13 was found in pollen formation. Also in 
Cornus glabrata, anaphase groups of 11 and 12 were observed. 
Nawaschin (1911) found in the meiotic division, in the pollen 
mother cells of Tradescantia virginica that one pair of chromosomes 
in the heterotype division is left near the equator, where it forms a 
vacuole which later becomes included in one of the daughter 
cells. He finds it dividing later, so that two cells of a tetrad 
usually receive 11 chromosomes and two 11 + this body, which is 
called a chromatin nucleolus or heterochromosome. But the brief 
account leaves many points undetermined. In the pollen formation 
of Callitriche verna , Winge (1917) has described somewhat 
similar conditions. The diploid number of chromosomes is 16, and 
in the heterotype division 7 pairs are usually distributed regularly, 
while the other two frequently pass to positions in line with the 
equatorial plate but far from the spindle, where they form nuclei. 
