938 
Journal of Agricultural Research 
Vol. XXXI, No. 10 
and although the work was conducted with great care there is a 
possibility nevertheless that a few stray pollen grain's gained admis¬ 
sion to the‘bagged branches of the Simple Leaf plants. 
The 158 first-generation carpellate plants were grown in the hot¬ 
house during the winter months, and under these conditions several 
of them produced some staminate flowers; otherwise all were typically 
carpellate. The small amount of pollen which they produced was all 
used to make self-pollinations. As no staminate plants were used in 
the same house during the growth period, the chance of possible 
fertilization by pollen from a staminate plant was eliminated, „ It 
seems reasonable to assume, therefore, that any accidental pollination 
which might take place through carelessness would be a cross between 
two carpellate-type plants. 
The number of staminate flowers available as sources of pollen 
was very small, and consequently the number of selfings which could 
be made was small. No cross-pollinations were attempted. The 
total number of seeds obtained was 193, many of which were small and 
shriveled. When these were planted the results shown in Table III 
were obtained. 
able III .—Results from self-pollinating carpellate type inter sex 
Cross 
Seeds 
planted 
Number 
dying 
young 
Number 
of carpel¬ 
late 
Number 
of stami¬ 
nate 
M—4 self _____ 
66 
2 
6 
0 
S-8 self....-__-_ 
30 
0 
1 
0 
S—1 self _ -__ 
6 
0 
5 
0 
M-3 self........— 
25 
0 
6 
0 
A-lOself .....-___ 
66 
12 
28 
0 
Total..______ 
193 
14 
46 
0 
The very poor germination of the seed is a most disappointing 
feature of these results. There is evidently a marked difference 
between the strains in this respect, but whether this is due to inherent 
characters or to other causes is a question which can not be answered 
from the data available. But of those which did germinate and grow 
to the flowering stage all were carpellate. Of course, these results do 
not prove that the progeny obtained by selling a carpellate type 
intersex will all be oi the carpellate type, but so far as they go they 
substantiate the results obtained in the previous generation. 
Up to this point results obtained by selling carpellate intersex 
types have been presented, and now it might be well to consider the 
possibilities of self-pollinating a staminate-type intersex. By this is 
meant a plant which is typically staminate until after the first few 
flowers have opened and then begins to produce also some flowers of 
the carpellate type. If this can be done it may yield some valuable 
information concerning the inherent differences between the carpel¬ 
late and staminate types. For several reasons, such a plan holds 
forth relatively few chances of success. In the first place, the devel¬ 
opment of functional carpellate flowers on the staminate type inter¬ 
sex is much rarer than is the production of viable pollen by carpellate 
type intersexes; and, secondly, the inherent difference in the length 
of life of the two types makes it difficult to coax the staminate plant 
