Water-Conductivity of the Wood in Trees and Shrubs. 247 



when it began to fall again. Doubtless the difference between Hartig's 

 observations and my own in this respect finds its explanation in the relative 

 severity of the German winter, which checks root absorption during that 

 period in the situations in which Birch is prevalent. This view seems to be 

 borne out by Hartig's statements concerning the Beech. He found that the 

 period of maximum water-content iu this tree coincides with the month of 

 December, and not with March as in the Birch. The Beech grows in less 

 exposed situations, and its close canopy likewise serves to mitigate the 

 effects of freezing winds on the soil. It is also of interest to note that the 

 Scots Pine in the Central European forests behaves like the Beech, although, 

 as might have been expected, the seasonal fluctuations in the water-content 

 are less marked than in the deciduous tree. It is evident, however, that it 

 is unjustifiable to assume that the results obtained under the influence of a 

 continental climate (e.g., of Central Europe, or many parts of North America) 

 must necessarily be identical with those yielded under such widely different 

 climatic conditions as prevail over the greater part of the British Isles. 



In the subjoined Table a summary is given of the investigations on the 

 greater number of deciduous species which have been studied in connection 

 with this research. I have not included a number of results on fruit trees, 

 because they form a special part of the general enquiry which is not yet 

 sufficiently ripe for publication. No attempt has been made in this, or the 

 preceding paper on the Evergreens, to deal with the range of fluctuation in 

 accordance with any fixed statistical principles, but a rather arbitrarily 

 limited range for each species has been adopted where experience seemed to 

 indicate the limits of normal fluctuation should be drawn. It might be 

 argued that it is useless to attempt to fix any limits other than those 

 afforded by the actual figures, on such a variable structure as that concerned 

 in water-conduction. But at least it is useful to recognise the relative 

 density or frequency of occurrence of the values within the extreme limits of 

 variability, and it is obvious that there is in most cases a genuine average 

 value which can be assigned for each species. It is equally evident, 

 however, that the proportional fluctuation is by no means identical for the 

 different species. Sometimes a reason for such fluctuation can be assigned, 

 as when heavily shaded (and consequently starved) shoots have been included 

 in the averages. 



It is not proposed at the present time to attempt a detailed discussion of 

 the results presented in the foregoing pages. To do so adequately would 

 require much space, and, moreover, there still remain many points on which 

 further information is desirable, before attempting to review the whole 

 subject. It may, however, not be out of place to state that there are good 



