219 
These percentages are not exceptional, having been repeatedly ob- 
tuined. I have not been able to secure germination in Chenopodium, a fact 
possibly due to the late collection of the material. 
Srconp.—In most cases the achenes show the highest germinating percentage vf col- 
lected at about the midd’e of the flowering season. 
If, for example, the flowering period is from July to October, achenes 
collected in the latter part of August or first of September will give a 
higher germinating per cent. than those collected at any other season. 
This is true in all forms studied, with the exception of Arctium. The 
non-viability of the later achenes may be explained as due to the action 
of frost. That of the earlier achenes is not so readily explained. In Arctium 
the achenes show a ready viability at all seasons. In the case of Bidens it 
is possible that the later achenes might show a fair percentage of viability, 
as no late collections of this genus were made. The difference in viability 
at different floral periods does not seem to exist in the other families 
studied. On the contrary, in the case of Abutilon and Solanum, the high- 
est percentages were obtained from the seeds collected very early in the 
season. 
Tutrp.—For the most part the central achenes of the head are not viable, and the 
same condition 1s frequently found in the outer rows. 
In all cases the largest germination percentage was obtained from 
achenes taken about midway between the center and periphery of the head. 
An exception to this statement is perhaps found in Helianthus, in which, 
in a single experiment, the central achenes were found to show a high 
germination percentage. A consideration of the order of maturing of the 
flowers in the composite head taken in connection with the mode of pol- 
lination furnishes the probable explanation of this fact. 
Fourtu.— The seedlings of all composite forms studied, were found to be particu- 
larly sensitive to both temperature and moisture changes. 
Very slight changes in either of these conditions, especially if sudden, 
proved fatal almost without exception. An increase in temperature of 
5° U., brought about in thirty minutes and continued for three hours, 
killed all the seedlings of Bidens, Cnicus, Lactuca, Solidago and Anthemis, 
the only composite seedlings escaping being those of Arctium. In the 
case of the other forms, Scrophularia alone succumbed, Abutilon, Malva, 
Solanum, Datura and Capsella not being visibly affected nor apparently 
retarded in their growth, . 
