Dixon and Atkins — Variations in Osmotic Pressure, Sfc. 237 



correspond with the occasional increases of the water in the soil due to rain- 

 fall. The lowest dip of the curve in November and October, 1910, may, in 

 this way, be attributed to the large rain-fall in those months (viz, 7"818 inches) . 

 The second dip (viz. in April) may be assigned to the slightly increased 

 rain-fall in March and April (2'08 inclies and 2'77 inches respectively). The 

 third dip at the beginning of July is hard to account for, unless we assume 

 an accidental local supply to the roots from a moderate raiu-fall on June 23 

 amounting to 0'370 inches. 



A few determinations on the sap of roots in February and March, 1910, 

 show quite different figures for the depressions of freezing-point. 



Table IV. 

 Ilex AquifuUum : iioots. 



They are much lower than those found in 1911. The observation on 

 March 5, 1910, is the lowest record for the sap of Ilex, viz. 0-294°, and 

 indeed for any plant we have so far examined, though in all close on 500 

 observations have now been made. We have found it futile to attempt to 

 correlate these figures with the rain-fall. In January and February, 1910, 

 the rain-fall was 5-826 inches, in March of the same year only 0-674. On 

 the whole it seems that the roots are able to maintain the cryoscopic values 

 of their sap irrespective of large variations in the amount of soil-water. 

 Possibly the great variations iu the osmotic pressures recorded for the roots 

 may be attributed in part to the fact that the roots dealt with were of 

 different ages. 



Owing to the difficulty of removing the undissolved substances from the 

 sap, only a few determinations of the mean molecular weight of the dis- 

 solved substances were possible. Filtration in most cases was impossible 

 owing to the viscid consistency of the sap ; or at least it took so long that 

 a change in the nature of the sap during filtration might be feared. At 



SOIENT. PROG. B.D.S., VOL. SUI., NO. XVm. 



2o 



