strous enough as to flowers, but absolutely without any flatness 
of stem, others slightly abnormal in both flowers and stem. Even 
the separation of those truly normal from the abnormals was 
difficult because true normals sometimes have flowers with an 
extra petal or two, due to environmental causes. 
When seeds from slightly abnormal plants were grown, some- 
times they bred true, givingriseto amodified race of abnormals; in 
other cases, seed of this type produced monstrous abnormal 
plants like the great grandparent and normal plants in the ratio 
of three of the former to one of the latter; while in still other 
cases, only normal plants resulted. Strictly normal plants always 
produced normal progeny so far as they have been tested, but 
there are grounds for believing that further tests will show certain 
normals will produce abnormals. The modified abnormals men- 
tioned above were only abnormal in their flowers, their stems being 
round and the number and arrangement of their leaves were also 
like normals. Monstrous plants resembling the 1907 ancestor in 
some cases produced only offspring like themselves, but in other 
cases their offspring consisted of various tvpes of abnormal 
plants. Figure 4 is a photograph of two offspring from a plant 
of the first generation of the normal x monster cross we have 
been talking about. Now what had happened to cause all this 
“messing up” of beautifully spun theories? Did the hereditary 
unit which differentiated the monster strain from the normal 
break up into numerous other units, or was it inherited intact 
and its ability to make itself known altered by the effect of other 
hereditary character-producing units, not encountered in the 
other normal strain. To me, the latter view seems the more rea- 
sonable, at least the more helpful, in trying to solve the problems 
of inheritance. So, upon this view, the difference in the results 
from crossing the monstrous strain with the two kinds of normals 
is explained by the difference in the kinds of hereditary units 
that made up the two normal races. In one case, the evidence 
leads us to believe that the monstrous and normal strain differ by 
a single hereditary unit, while in the case of the bushy normal, 
many hereditary differences were present, and some of these, such 
as that for branching, in expressing themselves, modified the ap- 
pearance of some of the plants that hereditarily should have re- 
sembled their monstrous grandparent. Of the other unit-differ- 
ences present in the bushy normal which helped to modify the ap- 
pearance of the abnormal characters, we know nothing as yet, 
for the unravelling of the heredity skein of even such simple 
organisms as tobacco plants means far more than the patient 
efforts of one man’s life time. For in order to study and isolate 
even one of these units, one must find a plant from which this 
particular unit is absent, and which in crosses with plants pos- 
