236 Scientific Proceedings, Royal Dublin Society. 



during the summer. During the months of November and December, 1910, 

 7-818 inches' rain-fall was recorded in Dublin. The soil must then have been 

 almost water-logged. Yet at this time the freezing-point depression of the 

 sap was in each case something over 0'900°. The rainfall for the six months 

 ending June 30, 1911, was 7"915 inches, but the depression of the freezing- 

 point of the saps was in each case less than 0'800°. 



In spite of the fluctuations, which are surprising, inasmuch as the above 

 conditions do not seem to exercise a direct influence upon them, the annual 

 curves seem to show two indistinct cusps best marked in the record of the 

 penultimate growth, one about November or December and the other about 

 March or April, with corresponding depressions, one about February and the 

 other about June and July. 



These depressions seem to correspond roughly with the ends of the periods 

 of elongation of the growths. In the autumn the buds may begin to open at 

 the beginning of October, and the axis may continue to elongate till January. 

 In the spring the leaves begin to unfold in May, and elongation proceeds till 

 July or longer. 



It is evident that during the elongation of the axis of the terminal bud 

 the various growths are at their youngest. Hence if concentration of the sap 

 proceeds with age, we may expect to find tlie smallest depressions of the 

 freezing-point of the saps coinciding with these periods of elongation. 



If all the shoots elongated simultaneously, we should get a sudden 

 depression in the curves corresponding with this elongation and followed by a 

 slow rise corresponding with the ageing of the growths formed. As a matter 

 of fact, however, the elongation of the shoots is by no means simultaneous, so 

 that the period of elongation is ill-defined, and consequently the depression is 

 gradual. 



In spite of fluctuations the curves traced by the freezing-points of the 

 two sets of leaves from the terminal growths nxn on the whole parallel to one 

 another (fig. 1). The two most marked divergences, viz., in October, 1910, and 

 June, 1911, are apparently attributable to the same want of simultaneity of 

 elongation. To take, for example, the divergence in June, 1911, here most 

 of the terminal buds had not elongated sufiiciently to have formed six mature 

 leaves. Consequently while the three ultimate leaves were for the most part 

 furnished by the buds opening in the spring, 1911, the three penultimate 

 leaves were chiefly furnished by growths of the autumn of 1910 which had 

 so far not yet begun to elongate their spring buds. Thus the ultimate leaves 

 were only just formed, while the penultimate leaves tested were mostly six 

 months old. 



With regard to the graph of the roots it might appear that the depressions 



