360 BOTANICAL GAZETTE [NOVEMBER 
1900, 6.81; April 21-22, 1898, 6.76; May 7, 1898, 6.51. A com- 
parison of these results with the curves in jig. 7 will show them to 
be strictly comparable with the conditions exhibited by Aster prenan- 
thoides in the single season of 1903. They differ, however, in being 
much less striking, the greatest change of mean value in Anemoile 
nemorosa being only 4.4 per cent., while the greatest change in mean 
value in Aster prenanthoides was 26.4 per cent. 
The Clifton area of Aster prenanthoides is in a perfectly natural 
~ condition, and though the region is much visited for its fine scenery, 
this particular spot, being less attractive to tourists and at the same 
time more difficult of access, is not likely to be at any time seriously 
disturbed. It can be assumed with perfect assurance that there 
were no appreciable differences in the habitat in the two years 1900 
and 1903, except such as were due to meteorological differences, and 
to these factors or possibly to internal periodicity, or a combination 
of these internal causes and climatic changes must be attributed the 
great differences found. 
It has not been infrequent to find great differences in variable 
characters of plants from markedly different habitats, as in the daisies 
(Chrysanthemum Leucanthemum and C. segetwm) collected from barren 
hills and fertile valleys by Lupwic and DE VriEs. But here at Clifton, 
Ohio, in the same spot, in the very same group of plants, undoubtedly 
consisting largely of the uniparental offspring of the very same 
individuals, the mean number of bracts was nearly 12.4 pet cent. 
less in 1903 than in 1900, the mean number of rays was nearly 10 
per cent. less, and the mean number of disk-florets 10.6 per cent es 
If such differences as these are due to climatic fluctuations, it ia 
of interest to consider what factors may have been important 5 
producing them. As already mentioned, REINGHL (1903) considers 
the chief factor in determining the number of parts in the androecium 
of Alsine media to be the condition of the available food-supP!) 
whether this be dependent upon the character of the soil or upom 
photosynthetic activity conditioned by the intensity of the sti 
As the physical and chemical conditions of the soil in the _ 
ravine were doubtless essentially the same in the two years 1D — 
the only soil factor which need be taken into account is water-SupP!Y 
as influenced by precipitation. REINOHL (1903) states that he co 
