J. A. Hahhis 
483 
Problem 4. Is there a ditference in the proportional number of locules with 
" odd " numbers of ovules in the eliminated ovaries and those which develop to 
maturity ? 
Locules with 4, 6, 8, 10 or 12 ovules may be conveniently designated as "even" 
while those with 5, 7, 9 or 11 ovules may be described as " odd." 
With respect to this character the fruits fall into four classes : 
3 " even," 
2 " even " + 1 " odd," 
1 "even" + 2 "<.dd," 
3 "odd." 
To the reader it may at first seem to be mere juggling with numbers to seek 
to determine whether there is any selective elimination of fruits depending upon 
the number of " odd " or " even " locules per fruit. Nevertheless purely embryo- 
logical considerations were the incentive to the work. 
I have not been able to find any account of the development of the ovary of 
Staphylea, but from the appearance of the matured organ and from our knowledge 
of the embryology of other forms it would seem quite likely that the ovules of 
a locule belong to the two margins of the same carpel. If this is the case, locules 
with "odd" numbers of ovules must have a different number produced on the two 
margins, while those with an even number almost as surely have the same 
number on the two carpellary margins. If this be the case no one will deny that 
structurally " even " locules are more perfectly formed than the " odd," being 
bilaterally symmetrical with respect to the number of ovules found on the two 
margins. 
The morphology of the fruit can be thoroughly understood only after an 
embryological investigation, but in the meantime the sensible thing to do seems 
to be to investigate the question of selective elimination with respect to this point 
on the basis of the material in hand. 
A reason for investigating this point lying within our material itself is the 
different frequency of fruits with " even " and " odd " locules. For the totals of 
the three collections the number of fruits belonging to each class is seen in our 
Table XIX. The predominance of locules with an even number of ovules 
suggests that this type may represent a point of morphological stability*. If such 
be the case the determination of whether there is a tendency towards the 
elimination of those with the larger number of " odd " locules is obviously 
important. 
The test may be made in two ways : (a) by the comparison of the percentage 
of "odd" locules in the three series, and (6) by calculating the percentage of each 
of the four different ovule formulae in the three collections. 
* Compare Sir Francis Galton's suggestion concerning points of organic stability in natural 
inheritance. 
