Ye} 
able, the flowers being long past. Inasmuch as there are several 
glomerules and since the plants are about 50 cm. tall, the second 
pair of alternatives in the key (p. ___) is readily reached but one 
tg unable to anawer either of these satisfactorily. An assumption 
is therefore made that the plant belongs in the first group. On ex- 
pripatios of the descriotions, however, it is found that the flowering 
scapes of those species are naked, but that this plant has a whorl of 
leaves near the middle of the scape. Passing to the next alternative 
group one finds that E. heracleoides satisfies this requirement and 
also that the plant corresponds to the description of that species. 
Thus by 2 series of assumptions and 4 process of elimination, the keys 
may be made useful in many cases where full data are not at hand. 
VARIATION: 
The effect of the environment may be seen in various ways. 
The physical composition of the soil and its consequent water-retain- 
ing power may greatly affect the stature of annual plants. -The effect 
of snow-lodgement may affect the habit of shrubs. The effect of wind 
on exposed ridges may affect the stature and habit of shrubs and trees. 
individuals growing in shade at the edge of the forest may have much 
larger and thinner leaves than those growing nearby in the open. Sucnr 
a list of ecologic and environmental differences or states mish*, be 
extended much further. To weigh correctly the valve of such variations 
requires care7ul observations and field experience. It is naturel tnat 
= + - ye to ~ , se ° = 2 - 2 : ee 2+} 22 rn 
the beginner is prone to overestimate variations which with wider ex- 
