6 9 
certain T ropica l T rees. 
from exposure to the sun so much more readily than the 
former — it must be inferred that the tougher, more leathery, 
mature leaves can bear with impunity a loss of water far 
greater than that which suffices to damage the thin, delicate 
immature leaves. 
The next step was to determine how far this relative 
lowness of transpiration on the part of the red leaves was 
due to their mutual overlapping and vertical downward 
position. 
The results given below show a difference— not however 
so marked as might be expected — between the rates of 
transpiration of the leaflets in the hanging overlapping 
(natural) position and when outspread. 
The smallness of the difference is probably to be explained 
by the fact that the experiment was made in a very moist 
atmosphere and in a room out of direct sunlight. 
The experiment was made by taking a healthy hanging 
branch of Amherstia nobilis , bearing a large pinnate leaf with 
red tender leaflets, and measuring the rate of transpiration, 
first, whilst the branch and the leaf hung vertically downward, 
and then when each leaflet was outspread in various con- 
venient positions. The leaflets were spread out by means 
of papers folded over stretched strings appropriately placed 
and gummed to the extremities of the long leaf-tips. 
The following are the results : 
Branch held obliquely up, but the leaf and 
Time in seconds of passage 
of bubbles up capillary < 
tube 
Time in seconds of passage 
of ten bubbles 
*5 
15 
14 
15 
15 
14 
14 
H 
1 5 
\i4 
*45 
Average =139. 
Leaflets spread out in 
. overlap. 
overlapping. 
14 
16 
16 
12 
14 
H 
13 
H 
12 
14 
12 
15 
14 
12 
J 4 
12 
13 
13 
13 
13 
14 
13 
14 
13 
13 
12 
14 
13 
13 
H 
15 
13 
13 
12 
H 
H 
12 
13 
12 
13 
13 
14 
H 
13 
14 
*5 
15 
14 
13 
14 
135 
138 
138 
124 
137 
Average =130-5. 
