SHULL: DUPLICATION .OF A ‘LEAF-LOBE FACTOR 429 
Australia, Tasmania, India, Ceylon, South Africa, the Sahara, and 
from widely distributed points in Europe and North America, and find 
that the forms everywhere fall into one or more of these four rosette 
types. This does not mean, however, that with respect to leaf-form 
there are only four biotypes of this species in existence, for nearly 
every lot of material from a new locality presents minor details of 
lobing which lead to their easy recognition as new and distinct biotypes. 
In all of the earlier crosses between types respectively dominant 
and recessive for either of the above-mentioned character-pairs, there 
appeared in the F,2 close approximations to the monohybrid ratio, 
3:1, or undoubted modifications of that ratio,—the modifications 
being due, in most cases at least, to the facts (a) that the A factor has 
been in some combinations not completely dominant, and ()) that 
both A and B require for their manifestation a certain minimum oppor- 
tunity in the way of favorable environment, including cultural treat- 
ment. 
These results having been well established while the situation in 
regard to the capsules called for further extensive investigation, a 
number of my cultures which concurrently involved the rosette char- 
acters and the capsule characters, have been grown under conditions 
not ideal for the development of the leaf lobes, though adequate for 
the determination of capsule form. For this reason my records with 
respect to the leaf types in certain families are of such incompleteness 
as to make the recorded ratios of no particular value. In nearly all 
cases, however, a small portion of each pedigree has been given suf- 
ficiently good treatment that the composition of the several families 
with respect to the rosettes could be inferred with small probability 
of error. 
The discovery to be detailed below, that in certain races there are 
two independent Mendelian factors which affect the leaf-form in 
identical ways, each dividing the leaf to the midrib and bringing out 
the secondary lobing which is seen unmodified in rhomboidea and modi- 
fied by the action of the A factor in the case of heteris, has revived my 
interest in the inheritance of the rosette characters and investigations 
are now in progress which I hope will give in time a full insight into 
the composition of the rosettes with respect to the major factors affect- 
ing the leaf lobes. 
The duplication of the gene which produces the triangular capsule 
has been found almost universally distributed geographically, as will 
be shown in detail in a later report, but the duplication of the leaf- 
lobe factor, B, appears to be relatively much less frequent. 
Before presenting the evidence of dimery in respect to the B 
lobes of the leaves, it will be advantageous to have before us the 
29 
