222 Prof. J. B. Farmer. On the Quantitative Differences in the 



The negative pressure is obtained by connecting the horizontal tube with the filter 

 pump B through the taps O and K. Any air bubbles that may be formed are easily 

 washed out by temporarily admitting water either through H or M. Practice enables 

 one to judge how much water is needed in the system to correspond with any desired 

 pressure, and, having once obtained this amount, the tap is turned to direct part of 

 the water into the burette P ; finally, any additional water required to bring the 

 water-level therein to the zero mark is di-awn in through K. Only one twig can, of 

 course, be connected up during a single experiment, and care must be taken that the 

 height of the water column on the scale G is always brought to the same level when 

 readings of the burette are to be taken. 



In practice, the results obtained under negative pressure merely confirm those more 

 easily and economically (since four twigs can be used simultaneously) reached under 

 positive pressure. This fact indicates that, as might be expected, the water is trans- 

 mitted with equal facility from base to apex and in the reverse direction, having regard 

 to differences of thickness of the two ends. It serves as a useful check on results 

 where there is reason to suspect air bubbles as the cause of unexpected irregularities in 

 consecutive readings, for the alternate application of positive and negative pressure may 

 be used to dislodge them without otherwise interfering with the experiment. At other 

 times also it afi'ords a useful means of testing a reading of unanticipated value. 



The subjoined Table indicates how the records were made, and the degree of 

 uniformity to be expected. It may be noted that the doubled pressure (60 cm. 

 mercury) is only applied for 1\ minutes, that being equivalent to 30 cm. for 15 minutes. 

 The readings are taken to the nearest tenth of a cubic centimetre in well-made measuring 

 glasses, graduated in cubic centimetres and tenths of a cubic centimetre. The stem in 

 question happens to be that of a deciduous tree, but it has an advantage in this respect 

 over that of an evergreen, in that the larger actual values given afford a better illustra- 

 tion of the method as a whole. 



Table I.— Apple, Cox's Orange Pippin, September 27, 1917. Length of 



Stems, 15 cm. 



Tin*. 



Stem 

 No. 1. 



Stem 

 No. 2. 



Stem 

 No. 3. 



Stem 

 No. 4. 



Pressure. 



4.3 to 4.18 P.M 



11 -4 



10 -6 



7-6 



4-8 



+ 30 cm. mercury 



4.19i to 4.27 P.M 



11 -2 



10 -6 



7-2 



4-6 



+ 60 cm. „ 



4.28 to 4.43 P.M 



11 -6 



10 -6 



7-4 



4-8 



-t-30 cm. „ 



Average absolute volume 



11 -4 



10 -6 



7-4 



4-8 





Area of wood in middle zone of 



2-68 



2-91 



2-06 



1 -47 





tvrig (Planimeter reading) 













The area of the wood is given in inches, as measured by the planimeter on a drawing to scale 

 ( X 10) from a section. It is reduced to cm. at magn. ( x 1) in calculating the specific volume. 



Perhaps the most clearly defined result which has emerged from this 

 investigation is that the evergreens, as a class, possess wood of markedly 

 lower efficiency than that of the broad-leaved deciduous trees. In other 

 words the specific conductivity is lower. Furthermore the absolute range of 



