ae 
78 MISSOURI BOTANICAL GARDEN. 
The differences between the two series are of precisely 
the kind which one would expect from the errors of random 
sampling, and from slight seasonal influences. The con- 
stants are chiefly of importance as a basis for the calcula- 
tion of correlations. It is interesting, however, to note for 
comparison with other characters, to be discussed later, that 
the variability in neither of these cases is at all large. Com- 
pare the coefficients of variation with those available for 
other inflorescences, in Table II. It will be shown that 
variation in both the number of seeds per fruit and the 
weight of the seeds is very much higher in C. longifolium 
than in most other species. This is probably referable to 
the large number of primordia formed and the large size 
attained by the seeds. ‘Taken together, these render the 
development of a high proportion of the ovules impossible. 
Some of the fruits produce a very large number of seeds 
while others tend to be sterile. Some of the seeds attain 
a very large size, but the majority must remain small. The 
immediate interest of these points lies in the fact that no 
influence of the peculiarities of seed formation upon the 
type or variability of the parts of the inflorescence can be 
detected. 
TABLE II. 
RELATIVE VARIABILITY IN INFLORESCENCES. 
mae Authe rit en d Coefficient of Variation 
phe aaeen Flowers Fruits 
Cichorium Intybus..... De Helguero, Biometrika. 5 : 185. 10. 28 — 10. 8. ie 
Staphylea trifolia...... Harris, Biometrika. 6 : 441. 64.4 53.04 
ee. ee Ritter, Beih. Bot. Centralbl. 25 : 21. 7 83 — 20. Te [xgeds = 
ee, Moschatellina . Harris, Biometrika. 7 : 220. 95 — 14.42 |...... 
Nothoscordum striatum Harris, Rep. Mo. Bot. Gard. 20 : 108. os 44— 25.51 |...... 
Allium stellatum....... 20 : 112. ae) Se 
Celastrus scandens ..... id Ree ee: * 20: 118. 26.80 48.58 
Cercis canadensis....... Harris, unpublished data. 12.09 — 24.97 |...... 
Sanguisorba officinalis. .| Ritter, Beih. Bot. Centralbl. 25 : 7-8. a 4 — 39.96 |...... 
Anethum graveolens ...} De Vries, The Mut. Theory. 1 : 558-561. — 26.50 |...... 
Coriandrum Sativum... me pee i 2 i oei. ia: 37 =e PEAR [iva ss 
2. The Distribution of Fertility in the Fruits. 
The fertility of a fruit is measured by the number of seeds 
which it produces. Table III gives the number per fruit 
and the frequencies for:— 
