120 BOTANICAL GAZETTE [aucust 
gether in plants with a low normal rate, as in the case of the geranium, 
A close scrutiny of Table I is sufficient to demonstrate that although 
the potometer will give some idea of the purely relative rates in different _ 
plants, it cannot be depended upon to give results which are at all 
comparable with their normal absolute transpiration rates. Thus 
if we arranged the plants according to their normal transpiration— 
rates, they would stand from highest to lowest thus: daisy, coleus, 
portulaca, geranium; but if a similar arrangement were made from _ 
the results of the potometer experiment they would fall into the fol- 
lowing order: daisy, geranium, portulaca, coleus. It was necessary x 
therefore to find some other method for measuring transpiration 
whereby the plant could be kept on its own roots in the soil and inas 
nearly normal condition as possible. 
As at least a partial solution of this problem, I would suggest the 
following method for measuring directly the transpiration of plants 
on their own roots, a method which provides at the same time for* 
supplying them with a constant and uniform current of air. The 
method is based upon the well-known affinity of phosphorus pentoxid 
for water, whereby the two are combined and phosphoric acid formed. 
This compound has long been used as a drying agent and as a meals - 
of separating from air or other gases their water-vapor content, for the 
purpose of measurement. To this end senha quantities of air are me 
drawn through U-tubes containing P, 0. oe : 
Fig. 1 will show the method of setting up the apparatus. The o 
apparatus consists essentially of a glass cylinder of suitable size to be 3 
used as a transpiration chamber, two U-tubes for P,O,, and a 
aspirator of known capacity. These are connected by rubber tubing, — 
so that as the water flows from the aspirator a known quantity of ait ait 
may be drawn through the cylinder and the pentoxid tubes. The 2 
cork in the top of the transpiration cylinder has two holes, one for oF 
the insertion of a thermometer and the other for the connecting tube 
to the pentoxid series. The cork for the bottom of the cylinder # has 
two holes; through one is passed a short bent tube serving as an 
for the outside air; the other is to accommodate the stem of the plant. 
The cork is cut in halves, so that it may be fitted around the pla 
before inserting i in the cylinder. If the stem be not large enough - 
the hole in the cork completely, this may be made. close by pack! 
