448 BOTANICAL GAZETTE [DECEMBER 
epidermis; the amount due to this cause was never found to be 
more than ro per cent. of the whole difference in thickness, and 
sometimes the epidermis of the new leaf was as thick as that of the 
old one. All the measurements were carefully made with the eye- 
piece micrometer. 
RELATIVE TRANSPIRATION OF OLD AND NEW LEAVES. 
The measurements of transpiration were all made by determina- 
tion of loss of weight of twigs in water, except in the case of Nerium 
of which leaves only were used, with the petioles immersed. The 
time allowed was usually one hour, and the temperature (always the 
same for each comparison) ranged on different days from 25 to 31° C. 
Observations were begun on May 8 and continued at intervals until 
August 6. : | 
In the following table are given the ratios of rates of transpiration 
for equal areas of old leaves and new ones which had just attained their 
maximum area. The values given are usually averages of several 
observations taken daily or at intervals of two or three days during 
_ a period of a week or ten days. The column o+m represents the 
ratio, amount of transpiration per hour for 100%™ of old leaves 
divided by the corresponding amount for new leaves. : 
TRANSPIRATION RATIOS; EQUAL AREAS. 
Olea europaea fre reg ne See <e 
Quercus Ilex - . - - - - 
Rhamnus Alaternus - - - - - 
Nerium Oleander — - - “ - - 
Pistacia Lentiscus ag oe eee kre - 
b O FH NOW FS 
in 
Fledera Helix - i ° = z : 18 
Smilax aspera - . i i . 16 
Viburnum Tinus = ‘ a = - 2.10 
It is obvious at a glance that in general the new leaves have not 
attained their full power of transpiration when they have reached 
the area of maturity. This fact is not wholly a result of the imperfect 
development in. thickness already mentioned, for when the transpira- 
tion ratios are referred, not to equal areas of leaves, but to equal 
weights, the older leaves often maintain their position as the more 
active, as is sufficiently shown by the following values. ‘These results 
cannot be closely compared with those of the preceding table, since 
