85 
after further drying at room temperature was preserved in tightly 
corked bottles. Extracts of this sifted soil in distilled water, in 
proportions varying from two parts to forty parts of soil per hun- 
dred of water were made, the water being allowed to remain four 
days upon the soil at room temperature, with frequent shaking. 
Similar extracts of 2zstfted soil from the two localities were also 
prepared. The filtered extracts were placed in chemically clean 
beakers, 400 c.c. in each. Lupine seeds were soaked in water 
six hours and then planted in moss. When the hypocotyls had 
grown to a length of about a centimeter they were marked five 
millimeters from the tip with India ink. They were then sus- 
pended so that their tips were just immersed inthe solution. To 
do this, glass rods were bent into L forms and pointed at their 
lower ends, which were thrust through the cotyledons of the 
a while their upper ends were pushed through oo 
sheet of cork which rested on the top of the beake 
ae an interval of eighteen hours each seedling was ae 
the root was gently dried with filter paper and placed upon glass 
over a ruler, where its length above five millimeters showed the 
growth for the period it had been in the solut Measure- 
ments were made at intervals of twenty-four ees ae four days. 
Seedlings of corn were used in the same way. 
In September, 1906, thirteen hemlock seedlings were placed 
in a pot containing soil from beneath the hemlocks and the same 
number were put in a pot containing soil from beneath the decid- 
uous trees. These were left in the propagating house of the 
w York Botanical Garden. At the end of three months all 
were alive and in good condition, though scarcely any growth 
was perceptible. After six months four of the seedlings in the 
emlock soil and five in the deciduous woods soil had died. 
Meantime the pots became choked with moss and liverwort and 
t of a year, September, 1907, only four survived in 
each pot. After a year and a half three were alive in each. 
This result, together with the irregularity of the curves that might 
be plotted from the tabulations given above, indicates that the 
failure of hemlock seedlings to germinate beneath the adult trees 
is not due to a special toxic constituent of the soil, but rather to 
