553 
Gates.—The Trisomie Mutations of Oenothera. 
have included approximately 50 per cent, of lata and 25 per cent, of 
Lamarckiana, the remainder being aberrant forms. Of the latter those 
known to have 15 chromosomes belonged to seven types. Only four mutants 
were known to have 14 chromosomes, and it is believed there were no others 
in this group. It appears from the rather involved account that the 18 
unclassified forms were believed to have probably had more than 14 chromo¬ 
somes. From lata x Lamarckiana Miss Lutz obtained 27 lata, 15 Lamarck¬ 
iana , 1 albida , 1 aberrans, and one doubtful form. From Lamarckiana x 
lata she succeeded in getting 1 lata, 63 Lamarckiana, 2 bipartita (?), 1 
nanella, and 9 of the rubrinervis type with 14 chromosomes which is prob¬ 
ably snbrobiLsta (see below). 
From all these and many other rather jumbled facts, several significant 
points seem to emerge as regards these trisomie forms : (1) scintillans gives 
a large percentage of oblonga offspring and evidently stands in a peculiar 
relation to that type ; (2) there is a reciprocal relationship between lata and 
scintillans in that each can give rise to the other ; (3) lata can produce 
directly albida, oblonga, and scintillans, while oblonga can produce albida 
but apparently not lata or scintillans, and scintillans produces oblonga and 
lata with relatively high frequency but apparently not albida ; (4) that semi- 
lata, Gates,' stands in a peculiar relation to lata has already been pointed out. 
We may next refer to the fact that albida and oblonga apparently breed 
true notwithstanding their unbalanced chromosome number. In the case of 
albida (de Vries, 1909 ) this involves only 122 offspring grown in 1897-8. 
At any rate no Lamarckiana appeared among them. For oblonga the 
evidence is better. Some 2,919 offspring were grown during 13 years, 
and were all oblonga except three albida, one elliptica,-ax\& seven rubrinervis. 
Miss Lutz ( 1917 ) suggests that such trisomie forms would breed true if all 
the gametes of one sex contain seven chromosomes and all those of the other 
sex eight chromosomes. In lata and other 15-chromosome forms it is clear 
that very few 8-chromosome pollen grains ever function in fertilization. In 
some cases, as I have previously suggested, the extra chromosome may be 
lost during the mitoses in pollen grain or pollen tube. It would therefore 
be necessary to assume that in albida and oblonga all the functional eggs con¬ 
tained eight chromosomes and all the functional pollen seven chromosomes. 
On the other hand, Bartlett ( 1915 ) probably overstates the case in assuming 
that all 8-chromosome pollen grains are eliminated, for various results indi¬ 
cate that trisomie forms do occasionally produce functional 8-chromosome 
pollen grains. As already pointed out in this paper, the frequent occurrence 
of trisomie mutants in pairs (de Vries, 1916 , also notes such cases) shows 
that these have probably arisen from non-disjunction in a pollen mother¬ 
cell of the previous generation. 
An effort has been made to systematize the many records of trisomie 
mutants in Oenothera as regards their origin, their offspring, and their 
