346 BOTANICAL GAZETTE [ NOVEMBER 
be as light as possible and perfectly centered, have a common 
axis, whose ends are slender cylinders, rolling on these plate-glass 
supports. The wheels are six and twelve inches in diameter, 
and absolutely true; when not loaded they are at rest anywhere 
on the supports. Over one of the wheels—the smaller one, as 
I have used the apparatus—runs a thread or string, which carries — 
on one end the plant whose transpiration is to be tested. The 
other end is fastened to an areometer, weighted until it is partly 
submerged. For this areometer I use a bottle partly filled with 
mercury, with a tight-fitting cork, into which is sealed a glass 
rod or tube. It is convenient to use a tube, so that the load of 
mercury can be adjusted without. disturbing the cork. For the 
string I used double heavy silk thread, boiled in beeswax, and 
rubbed until it would not stick to the wheel. It was fastened to 
the upper end of the glass tube, out of contact with liquid water, 
and altogether was pretty well protected against hygroscopic — 
changes in length. 
Now, as the plant transpires it becomes lighter, and the 
areometer sinks, displacing exactly the mass of water at that 
end of the string which has been lost by evaporation at the 
other. If, for example, the area of a section of the tube be 
i, it will sink 1™ while the plant loses 1% in weight. The 
larger wheel carries a thread, with a tracer which leaves its 
record in the same way as that of an auxanometer. 
What are the limitations of the working of the machine ? 
When it is used with proper care, there is practically but one, 
the inertia of the resting load which the wheel carries. Friction 
is practically eliminated. The axis turns more easily than would 
be feasible on ball bearings. The only remaining obstacle to 
perfect ease of movement is the surface tension of the waters 
but the capillarity is not very considerable even at its theoret- 
ical maximum, which is never reached, and if the tube is unr 
form and clean it will hardly vary as the tube descends. Jats 
and irregular drafts must of course be avoided. 
It has been possible to put the apparatus to an unfortunately 
brief test. Both potted plants and water cultures were used j 
E. 
a 
iy 
