1908] FREEMAN—DETERMINATION OF TRANSPIRATION 123 
‘ 
thermometer, or by passing a given quantity of the normal air at the 
time of the experiment over phosphorus pentoxid in U-tubes in the 
same manner as in the case of the transpiration experiment itself. I 
prefer the latter method, as it is more direct and accurate, corrections 
are not required for altitude, barometric pressure, etc., and it calls for 
little extra labor to set up the additional apparatus and to run it 
simultaneously with the other. If all of the aspirators be of the same 
size, it will simply be necessary to subtract the increase in weight of 
the tubes used for the normal air from the increase in weight of those 
used in the plant series to find the actual amount of water vapor 
transpired by the plant in the given time. By dividing this amount 
by the product of the area of the leaves and the duration of the 
experiment expressed in minutes, and then multiplying this quotient 
by 60, the transpiration rate per hour per unit area is obtained. 
; The degree of accuracy of this method may be measured by check- 
Ing it against the transpiration of a cut stem in a potometer or of a 
plant in a sealed pot. In my experiments I used the cut stem of an 
alfalfa plant. The leaves were removed from a few inches of the 
lower part of the stem, which was then inserted in a bottle of water 
through an opening in the cork which was thoroughly sealed over 
and around the stem with paraffin. By making several weighings of 
this plant before and after its use in the transpiration apparatus, its 
Tate of loss of water was determined for the time before, during, and 
after its inclosure in the cylinder. The net increase in weight in the 
P.O, tubes, after the weight of water present in the amount of normal 
“t used has been subtracted from their total increase, should of course 
equal the loss in weight of the potometer during its inclosure in the 
cylinder. This method of checking was repeated three times with 
Ae ease results, details of which are shown in Tables Il 
It will be noted from Table II that the last weighing of the potom- 
aid Was made at 3:10, or two minutes before it was placed in the 
fs linder, and that the first weighing thereafter was made at 4:18, or 
ve minutes after the air ceased to pass over the P,O, tubes. The 
ein of water transpired during these intervals must be subtracted 
om the total of 0.150 grams as shown in Table II before comparing 
“with the net gain of the P,O, tubes. Again referring to Table I, it 
