WAYFARING NOTES FROM THE TRANSVAAL. 58 
and { gives to the wind. After umeaat the tog increases Be: 
among them gladly fastens upon any feature that may aid him 
classifying, and it would seem that note of physiological futiotion 
may often be ogee perm taken in conjunction with considera- 
tion of structur 
In the few South African Composites seen by the writer, ex- 
cluding the more obvious characters, points which may aid woul 
st to be— 
. The hairs or emergences of the achene: their character, 
diapes disposition ; whether evenly neg tee over the surface of the 
achene, arranged in] row ,or ega 
at the base. The coronal distributive hairs (pappus | proper) lear arly 
function as the medium by which the wind spreads the fruit abroad, 
but the structure (often A apeae or hooked) of the hairs upon the 
body of the achene would appear to point to their being fixative, 
se distribution is aegis liche d. 
. The disposition of the barbs upon the pappus- race i.e. upon 
the. main body of the shank, whether closely set or placed at 
extended interval. This apart from oy closer ranging which is 
— towards the aa of the hai 
. The shape the corolla of the disc-floret merits close 
bomiptisn; and sip point at which the filaments stand free from 
the es ollal tube. 
- The bulging of the filament, so frequently seen below the 
nate is often characteristic. It may be clear when the rest of the 
filament is tinted, and the griaskauiniins of the surface-cells in that 
region is closer than upon the surface of the filament generally. 
e. Variations in the shape of the projecting tips of the anthers 
may well be noted; usually hastate, they may be cordate, or have 
their tips Shes e d off. 
-f. As to the significance of the ‘anther. tails.” Where they 
have no whisk and are merely dwindling prolongations of the lower 
ends of the loculi, - is hard to see that they have much functional 
value. The tails are almost of necessity absent where the dise- 
Hore has an itseclaeal tube 
. The pollen grains. Their colour, shape, size, quantity, and 
darldetherkings: 
. Dark colouration of the external face of the ray-floret, for 
this appears to me ee eee with exceptional sensitiveness to 
pent come chan 
. The depth se ‘the involucral cup given in millimetres. 
. The involuere, of course, in many cases, actively participates 
in ihe closing of the flower, and its description might be made with 
this kept in view. 
How great » variation in_colour may take place. sie 
limits of srieteems may be seen in Gerbera discolor Sond. and G, planta- 
ginea Harv. (a small form). Bo oe an say may bo of pale 
