ECOLOGICAL ADAPTATIONS OF CERTAIN FERN PROTHALLIA 48 1 
instead of the funnel tube A-B and allowing the water and air to flow 
from the free end of the tube below. Water passing slowly from the 
tap through Q into the chamber carries air through the curve at 5 
and reduces the pressure in R so that a current is forced through P by 
external atmospheric pressure. When P is connected with the 
delivery tubes of the other pieces of apparatus the air current passes 
through them as before. Control of the air current is secured by 
screw cocks as above. The actual difference of pressure in the desic- 
cating tubes with the different methods of securing pressure is slight, 
and as results have shown, is wholly negligible for the work in hand. 
I. A glycerine desiccator was so arranged (text-fig. i) that a current 
of air passed first through a wash bottle, E, filled with crumpled filter 
paper saturated with glycerine, and then as bubbles of approximately 
.1 cc. capacity, through a column of pure glycerine I m. long in a 
Winkler's potash tube, F. The passage of a bubble through this 
tube required 10-12 min., the time variation being due to variations 
of temperature and the resulting difference in the viscosity of the 
glycerine. The dry air next passed through 12 cm. tubulated drying 
tubes, G, and finally out through a 250 cm. wash bottle, H, filled with 
fused calcium chloride. This last element was used to prevent any 
possible entrance of moisture by diffusion or by currents resulting from 
changes in temperature. The clumps of plants and soil were carefully 
removed from the culture saucers when the soil surface began to show 
dry, and placed in the drying tubes, G. They were thus directly in 
the path of the current of dry air. 
II. The second apparatus was an exact duplicate of that described 
under I, but the Winkler's tube contained aim. column of c.p. 
sulphuric acid, sp. g. 1.84. The air bubbles, of about .03 cc. capacity, 
passed through the acid in a little less than i min. In check experi- 
ments the presence of injurious gases was guarded against by placing 
the tube, /, in a jar, as shown, through which a continual stream of 
water passed, keeping the temperature of / at from 13° to 15° C. 
III. A third set of apparatus was so arranged that the current of 
air passed through a 100 cc. calcium chloride tube, L, and then 
through four U-tubes, M, all filled with medium size lumps of fused 
calcium chloride. This series of tubes was equivalent to a i m. 
column of the chloride. From the tubes the dry air passed over the 
specimens in the drying tubes, N, and out through the final check tube, 
0, filled with the chloride. To insure continual drying of the air the 
