34 
must be cut out after they have grown almost to maturity, resulting 
in unevenness and irregularity in the orchard and much loss of time 
and space. For this reason, together with the difficulties of breeding, 
the dioecious type probably will be largely eliminated. 
BREEDING THE HERMAPHRODITE FORMS. 
Turning to these forms with more hope of results, some experiments 
are being conducted. Here it is possible to deal with a single indi- 
vidual mother plant. One of the first facts to be determined is the 
extent to which the offspring of such a parent may be expected to be 
fruit-bearing trees. It was known from earlier observation, with- 
out any definite experiment, that a large number of the seeds from a 
fruit of a hermaphrodite flower, produce fruit-bearing trees, either 
pistillate or, like the parent, hermaphrodite. 
A tree was found in a Honolulu papaya orchard producing fruit of ex- 
cellent flavor. The fruit from which the seed was taken was of the long 
cylindrical form, but it can not be stated that all the fruits on the tree 
were of that shape. Its flowers were staminate and hermaphrodite, 
and so far as observed, of the elongata form. The flowers had not 
been hand-pollinated and it is therefore impossible to state whether 
they were autogamously fertilized or otherwise. The seeds were 
planted May 2, 1910, and later 35 of the young plants were set in the 
orchard. Of these, ¥ v 34 were hermaphrodite and 1 was a staminate 
tree. The hermaphrodite flowers on most of the trees were of two 
types, some of the elongata form and others resembling pentandria, 
with corresponding difference in the fruits. All of the 35 trees were 
designated as No. 2355, and, according to the method adopted in all 
our breeding work, the individuals were designated thus: 2355:1, 
2355:2, etc. The best of the trees from the standpoint of uni- 
formity of cylindrical shape in fruit was 2355:1, which was also of 
very good flavor and a reasonably good producer. This was selected 
for further breeding. Two of its flowers were hand-pollinated each 
with its own pollen and protected from all possible allogamy. The 
seeds collected from these two fruits were planted as No. 3198. 
At this writing there are of this F 2 , 343 trees living and old enough 
to have exhibited sex characters, some of them having done so for 
several weeks past. The individuals judged by the characters 
apparent to date are as follows: 
Pistillate trees 98 
Elongata 61 
Pentandria 45 
Elongata, pentandria, and intermediate 55 
Pentandria and pistillate 1 
Fruit bearing but not determinable in form 62 
Total fruiting exclusive of correse 322 
