Priestley and Pearsall . — Growth Studies. III. 487 
internal diameter of the tubes Z W was less than 2 mm. A tap at N allowed 
for the removal of air-bubbles. 
The further details of the apparatus can best be realized from a 
description of its use in measurement. A container holding a plant whose 
roots were to be measured, and filled with the nutrient solution supplied 
to it on the last occasion that the volume was measured, was clamped 
vertically in a stand. The connexion M, at the base of the measuring tube, 
was attached to the outlet A of the container. A pump connexion was then 
attached at Z, and the solution from the container sucked up the measuring 
tube as far as was necessary. The tap U was then closed and the pump ' 
disconnected. By adjusting U, the level of the fluid in the tube was then 
allowed to fall to Z, Y, or X as required. Tap C was then turned and the 
solution in the container allowed to run off through B, until the meniscus in 
the container was just level with the top of the tap outlet at C. Before 
making the final adjustment here, A was opened to allow the pressure in 
U M to become equal to that of the atmosphere, since otherwise it would 
vary with the height of the column in the measuring tube, and a volume 
error would be introduced. When C was adjusted the connexion on B was 
attached to P, no air being included. C was then opened again and the 
required volume of fresh nutrient solution run in from the bulbs, by 
adjusting tap Q. B was then closed and A opened, when, by adjusting tap U, 
the final small amount of fluid was run in until the surface of liquid in the 
container just touched the porcelain point in the neck. The reading on 
scale w X then gave the volume added from the measuring tube. As the 
roots of the plant grow, this becomes less, and the decrease in this reading 
indicates the volume increase in the roots. If there is any doubt as to the 
accuracy with which this end-point has been measured, the solution can be 
sucked up into the measuring tube and the final operation repeated. 
The sources of error in this apparatus are obviously due chiefly to the 
number of readings of meniscus level. In any series of observations with 
the same container, however, these readings are made at the same six 
places in the apparatus, so that this source of error is continuous throughout 
the readings. Since all of these levels, except that at D, are observed in 
fixed and narrow tubes, they should permit of accurate determination. In 
the mouth of this container, however, the accuracy of the end-point in filling 
must depend largely on whether the container is fixed in an exactly vertical 
position or not. In our containers, the body was cylindrical, and its correct 
position was ensured by getting the edge in exact alinement with the 
upright of a stand behind it. 
Air bubbles in the apparatus are another source of error, but they are 
easily seen and as easily avoided. For obvious reasons, the temperature 
of the apparatus should be kept approximately constant at the different 
readings. 
