104 VEGETABLE PHYSIOLOGY 



its position can be removed by causing it to collect at the 

 upper portion of the straight tube of the potometer. To 

 take an observation of the rate of transpiration of the 

 branch, a bubble of air must be admitted into the capil- 

 lary tube by momentarily removing the vessel into which 

 it dips, and replacing it as soon as the transpiration has 

 caused the air to enter. The bubble of air must be of 

 uniform size in successive readings, to ensure that the latter 

 shall be strictly comparable with each other. The bubble 

 will rise in the tube, and finally make its way to the upper 

 part of the straight limb of the instrument, the rate 

 at which it travels serving as an index of the rate of the 

 transpiration. The capillary tube should be marked 

 by a transverse line a few millimetres from its lower end, 

 and by means of a stop-watch the time taken by the 

 bubble to rise from this mark to the free end of the tube 

 should be observed. The branch may be covered by a 

 bell-jar, so that the variations of temperature, moisture, 

 &c., of the air surrounding it can be controlled during a 

 series of observations. Less accurate observations can be 

 made by substituting for the capillary tube a tube of wider 

 bore bent at right angles a little below the orifice of the 

 potometer, and affixing to it a scale by means of which 

 the rate of passage of the column of water in the tube can 

 be observed (fig. 63). 



According to the variations in the external conditions of 

 the plant, including all the features already alluded to, the 

 amount of watery vapour transpired is continually changing. 

 The most favourable conditions being afforded in summer, 

 it is not to be wondered at that transpiration attains an 

 annual maximum during that season. It does not, however, 

 entirely cease during the winter, though it is reduced to a 

 minimum, especially in the case of such trees as shed their 

 leaves in the autumn. 



Apart from such changes in the external conditions, 

 transpiration appears to show no independent periodicity, 

 differing in this respect conspicuously from root-pressure. 



