217 
botanist knows that this plant makes no visible increase in numbers from 
year to year. The great number of capsules filled with seeds is but the 
sign of its intense struggle. 
In many cases the causes which hold in check the undue increase of 
a specific form are evident. The law requiring that stock should be kept 
within bounds, modified in very marked degree the flora of the State, and 
in certain regions has served to cover stripped hills with a new timber 
growth. But there are cases in which the number of seeds produced is 
so enormous, the means of dispersal so various and ingenious, and the dis- 
persai itself so sure, that we wonder that the increase is no more rapid. 
In this category we find notably the composites. Considering their myr- 
iads of seeds and their perfect means of seed dispersal, their increase in 
numbers is insignificant. Indeed, in many cases special protective devices 
have been developed in order that they may maintain their place in nature. 
It is evident that factors other than those ordinarily limiting plant 
distribution are operative in limiting the distribution of the composites, 
or that the composites are peculiarly sensitive to conditions which do not 
materially affect the majority of plants. The theory of special limiting 
factors seemed scarcely worthy of consideration. It was therefore pre- 
sumed that some, at least, of the ordinary factors were more effective than 
in other cases, and that in all probability there was a peculiar sensitive- 
ness to these changes at some particular stage of development. The ex- 
periments were undertaken with the hope of throwing some light upon 
these points. The conditions of the experiments have been varied from 
time to time as experience suggested, and to such an extent as to preclude 
tabulation in the limits of this paper. Conditions may be summarized as 
follows: Two kinds of soil were used—a loam mixed with sand and a 
leaf-mold. In both cases the soil was carefully sifted and packed in the 
pots in order to prevent subsequent settling. As regards moisture, three 
conditions were used—saturated, moderately moist and extremely dry. 
The percentage of water in the soil was not carefully worked out, but 
was roughly estimated to range from 25 per cent. in one extreme to 
between 80 and 90 per cent. in the other. Planting was either surface or 
at about the depth of the seed. The average temperature in the first 
series of experiments was 26.5° C., with extremes of 20° ©. and 31° C. In 
the second and third series the average temperature was 24° C. As far 
as could be determined by inspection only perfect seeds were used. As a 
matter of course, control experiments served as checks upon results. 
