226 MR. I. H. BURKILL ON SOME VARIATIONS IN 
Richness of the Soil—We might have almost expected that 
plants of B growing freely, removed from competition, would 
have produced more stamens than those of A, where the compe- 
tition was severe. But such was not the case. E, F, J, and K 
were, as already stated, grown—E and J in rich soil, F and K in 
soil one half of which was sand. The seeds were collected in 
1892 in the Botanic Garden, and sown in these four sets in 
February; the pots containing E and F were then placed in a 
hothouse, while J and K were grown out of doors. The following 
table gives the results obtained :— 
TABLE V. 
Percentage of Stamens and Staminodes in Flowers of Stellaria media 
grown in Rich and Poor Soil. 
No. of t °/o of MO 
flowers. |? Dnes. stamens 
present. present. 
E. Rich soil, hothouse, April - . . 
13 to June 23... | 251 2:19 37:83 
F. Poor ,, » 5 eI 101 4:16 96:34 
J. Rich soil in open, May 7 to . . 
June 30 ......... ] 360 36 4894 
K. Poor ,, » poene 90 :22 45:22 
Note.—J and K were accidentally killed ; otherwise the later flowers would 
have reduced the high percentage of stamens. These plants were slower in 
growth and more vigorous than those of E and F. 
The percentages of stamens in E and F and in J and K in 
these four experiments lie so very near to one another, that poor 
nourishment to the extent here produced seems to have had no 
distinct effect on these organs. The only decided influence which 
the sand in the soil seems to have had is on the number of flowers 
borne by the plants themselves. 
As soon as E and F had finished flowering, the pots were 
resown (July 6th), in order to see if the soil would bear freely a 
second crop. The plants of this second sowing began to flower, 
those in poor soil on August 2nd, those in rich soil on August 11th; 
but I was not able to observe them thoroughly. The new plants 
in both the poor and rich soil were strong; and of the thirteen 
