Harris, Observations on the Physiology of Seed Development in Staphyleci. 15 
Accordingly a difference of one node in tlie position of a fruit 
would be accompanied by no more change in tbo mean nnmber of 
seeds tban 1,100 to 2/100 seeds. This arnonnt is so small that 
witb no more tban two thousand fruits, one cannot even be sure 
of tbe sign of the relationship. Diagram 1 shows by the slope 
of the lines and the plotted empirical means, how insignificant the 
correlat.ion is. 
The mean of the three regression coefficients is only —.0035 
while for the regression of ovales on position as fonnd in another 
place we have the average value —.0933. This result agrees with 
the conclusion drawn front the individuals of the 1906 collection, 
that the number of seeds developing is less dependent upon the 
position of the fruit on the inflorescence axis than is the number 
of ovules formed. I believe this result will be surprising to many 
botanists. 
The same thing is made very clear by combining the three 
series of material and determining the coefficients of fecnndity for 
the first five nodes of the inflorescences. We have: 
Position on T , , a i 
T n total Ovules 
lnnorescence 
Total Seeds 
. Coefficient 
of Fecunditv 
1 
97420 
10224 
.1049' 
2 
66504 
6817 
.1025 
3 
23224 
2469 
.1063 
4 
3094 
345 
. 1115 
5 
334 
35 
. 1048 
V. Summary and conclusions. 
1. The foregoing pages contain the tabulated data and the 
results of analysis for three series comprising altogether over eight 
thousand fruits of Staphyleci trifolia. The purpose of the work is 
to ascertain something concerning the internal factors iniluencing 
the development of the seed. 
2. Wuch dogmatism prevails among biologists concerning the 
“explanation“ of “fluctuating variability“. It is frequently assumed 
that any Variation polygon based upon “pure“ material may be at 
once interpreted in terms of external environmental influences and 
internal dilferentiation. 
3. Where an individual produces a series of Organs which 
differ among themselves, it is generalJy assumed by biologists that 
differences in the vigor of individual branches or those due to 
“periodicity” will largely. account for the "partial variability’’ 
observed. In some cases the^e factors probably have a considerable 
influence on the determination of the characters of the Organs of 
an individual, but in the present material, the relative as well as 
the actual number of seeds developing seems to be very little 
