348 BOTANICAL GAZETTE [NOVEMBER 
numerically by the standard deviation 7.692 -+.133, or by the coeffi- 
cient of variability 19.928 +.345. 
2. Rays.—The ray-curve shown in jig. 9 has the mean at 25.247 
+.122 and by much the strongest mode at 22, so that here again 
there is a very prominent skewing of the curve. The empirical 
| 
5 EB 
it 
SEER GW mi 
(Be NRT a Bs He BE Se 
ie ee en 2 2 
) Sh Saal 2A Sd 
| 
i 
| 
iI 
a a 
sR 
| 
Fa SS a A 
OS: WR A OR LS Oo 
bei 
<t 
Bae See 
it 
, eG i RE Re ee al a 
20 25 30 carey 45 50 35 
- em ir- 
Fic. 8.—Bract-curve for 757 heads collected in 1903: mean 38.597+- 189; emp! 
ical modes 29-30, 34, 30, 41, 46, 49, 51; 7=7.692+ 
modes at 16, 22, 26, 33 are somewhat suggestive of the series ie 
Lupwic, DE Vrigs, and others have shown to be so common, Pee 
aside from such suggestion can have little meaning. The stand pe 
deviation 4.990+.087 is considerably less than that of the. ae 
while the coefficient of variability, 19.764--344) 1S practically 
same. 
