J. A. Harris 
455 
12 ovules are borne in two series along the inner angle of the locule. None, one 
or a few of the ovules in each cell or carpel develop into smooth, bony seeds. 
The flowers are borne in racemes or panicles. Only a small percentage of 
those formed in an inflorescence mature fruits. I am able to present no reliable 
statistics on this point, but I would say that in the individuals which I have 
studied not over 10 per cent, of the flowers of an inflorescence which ripens fruits 
at all mature their ovaries. What proportion of inflorescences fail entirely to ripen 
fruits I am quite unable to say. 
It has not seemed worth while to secure quantitative data on this point. The 
exact proportion varies from season to season. It is always so small that a very 
stringent elimination is obvious. 
This memoir is limited to two series of data, (1) that especially collected for the 
purpose in 1908, and (2) a small lot of ovaries in different stages of development 
taken for other purposes in the spring of 190G. 
1. The 1908 Collections. 
The collection of this material and the counting of the number of ovules per 
locule was carried out by my two assistants, Miss Eva N. Dixon and Miss Rose 
M. Peclimann, and my sister. Miss Nellie L. Harris. I regret not having been able 
to attend to the sampling personally, but the necessity for an investigation of the 
problem of selective elimination in Staphylea became apparent to me after I had 
left the Missouri Botanical Garden, and the necessity of devoting the spring 
season to experimental plantings at the Station for Experimental Evolution, 
precluded my visiting the Garden at the flowering season for Stcvphyleci to make 
the collections myself. 
As for the countings, I cannot state too highly my indebtedness to these 
assistants for the faithful way in which the arduous task was carried out. In 
all over 21,000 locules had to be opened and the number of ovules counted. For 
the mature fruits the opening was simple, but the counting of the aborted ovules 
is not always an easy task. For the immature ovaries the work was much more 
trying to the patience and eyesight. It was quite out of the question to deal 
with the fresh material, and the difficulty of dissection and counting was increased 
by preservation in alcohol. First the ovaries had to be dissected out of the 
flowers and then each of the locules opened with a needle and the number of 
ovules counted. Possibly some of my fellow botanists would have had the deftness 
of fingers and the patience to carry through the countings, but I would have found 
the task impossible, and even Miss Pechmann and Miss Dixon occasionally nearly 
lost heart during the several months the work was under way. 
The collection of material to determine whether or no there is a selective 
elimination is a difficult problem in a form like Staphylea. The number of 
flowers produced is very, large, and it is quite impossible to take account of 
