Vou. 4 
82 ANNALS OF THE MISSOURI BOTANICAL GARDEN 
Garden in March, 1915, one rootstock is of teratological in- 
terest. It produced three stems exactly alike. Іп every case 
the flower and peduncle were entirely absent, and there was 
a whorl of 6 rather small (about 5 em. long), nearly equal 
leaves. This rootstock has been marked and will be observed 
to determine whether the same abnormality occurs every year. 
It is evident that most of these scattered records were un- 
known to those who recorded their own observations. It is 
therefore useful to bring a number of them together, and no 
doubt this list can be considerably added to. Т. grandiflorum 
appears to be the most variable of all in certain localities, 
and it is obvious that in the different districts where studies 
have been made, much the same series of variations and ter- 
atological malformations have been encountered, though the 
forms with the stalked petals or ovary appear to be more 
restricted. It is proven that these are not environmentally 
produced, at least in the sense that their recurrence from the 
same root year after year is independent of environment. We 
can only suppose that such rootstocks have been produced 
from particular seeds in which a mutation had occurred, giv- 
ing rise to one of the many aberrant conditions found. The 
species is in an unstable condition in the same sense in which 
I have used that term for Oenothera Lamarckiana. It is pos- 
sible that cytological study of Т. grandiflorum might reveal 
the basis of this unstable condition, as it has done to some 
extent in Oenothera, and a careful study should be undertaken 
with this possibility in view. It seems evident that Т. grandi- 
florum is mutating in much the same sense that the term can 
be used for (Е. Lamarckiana. In T. grandiflorum, however, 
the mutations are for the most part teratological. It is im- 
portant to discover, if possible, the fundamental difference 
between the condition of the germ-plasm of T. grandiflorum 
in which the variations are chiefly in number and arrange- 
ment of parts, and the condition in Oenothera in which the 
variations are better coordinated — changes occurring simul- 
taneously in all parts usually without dislocation of their rela- 
tion to each other. It must be supposed that a redistribution 
