348 BOTANICAL GAZETTE [NOVEMBER 
In measuring the distances registered on the smoked cylinder 
I attempted accuracy to 0.1™", which is rather finer than is safe. 
I believe that the real limit to the accuracy of the results to be 
obtained with this apparatus lies in our ability to measure the 
trace. The weight on each side of the wheel in this test was 
only about 700%", less than would often occur in practical use. 
It has not been practicable to make a test with a greater load 
than 3.5"%, with which, under proper conditions, it responds to 
a change of not more than 50" in either direction. 
It is a merit of this apparatus that it will register equally well 
a decrease or an increase in weight, without any change in the set- 
ting up, except as the areometer will be set deep in the water if 
a continued increase is anticipated. It can be used to measure 
the changes in weight of fruits, etc.; and with some modification 
in details, the plan of the experiment is a good one for delicate 
and accurate measurements of the pressure of growing roots, 
the lifting power of prostrated grass stems, etc. 
Finally, one comment on the table, introduced only as an 
illustration, a further discussion being reserved for a future 
time : the remarkable transpiratory activity during a few hours 
of both forenoons is not an error, nor does it seem to be am 
accident ; for four different plants of red clover showed the 
same striking behavior, 
MADIsoN, WIs. 
