SOMATIC CHARACTERS AND FERTILITY 
409 
was determined by clearing with alcohol and counting under the mic- 
roscope, the results are^^ 
Tree i, r = - .007 zh .023, 
Tree 2, r = .030 =b .021, 
Tree 3, r = .134 =b .024, 
where the probable errors are calculated on the basis of the number oi 
ovaries dissected out and examined, not the number of inflorescences^ 
Thus the correlations between the number of flowers formed of 
number of fruits matured per inflorescence and the fertility of the 
fruits — either as measured by the number of ovules laid down or the 
number of seeds ripened — are without exception very low indeed. 
IV. Summary and Discussion 
This paper, which is one of a series on the various aspects of fer- 
tility and fecundity in plants, is a contribution of data towards the 
solution of the problem of the relationship between the degree of so- 
matic development of the plant individual as measured by the number 
of fruits which it bears, and the fertility of these fruits as measured by 
the number of ovules formed and the number of seeds developing. 
The data comprise the records of number of ovules formed and 
number of seeds matured in 127,610 pods of garden beans, Phaseolus 
vulgaris, from 19,064 plants with known number of pods drawn from 32 
cultures made under a wide range of environmental conditions and 
embracing five different varieties. 
Taken in connection with data recorded in other papers they per- 
mit the consideration of points not touched upon here. It is perhaps 
fair to state that these are to be discussed in subsequent papers where 
extensive additional series of records already in hand can be reduced. 
The analyses of the data as far as carried out lead to the folowing 
conclusions: 
(1) The correlation between number of pods per plant and number 
of ovules per pod has always been found positive but low, ranging from 
.023 to .355, with an average value of about rpo = .195. 
(2) For pods per plant and seeds per pod the correlations are also 
for the most part positive, although constants which have the negative 
sign but are insignificant with regard to their probable errors some- 
times occur. 
These values range from — .046 to .338 with a mean of r^^ = .126. 
^3 Bot. Gaz. 53: 403-404. 19 1 2. 
