UPWARD TRANSPORT OF NITROGEN 67 



estimated a loss of 10 per cent of the dry matter over night.) 

 Expressing nitrogen as a percentage of dry matter this 

 change in dry matter would cause a seeming gain of 

 nitrogen from a content of 5.25 per cent to one of 5.71 

 per cent, an apparent gain of 8.75 per cent. The order of 

 error, therefore, under the assumptions here made, seems 

 about the same whether expressed as percentage of dry 

 weight or of fresh weight, if, on the one hand, the change 

 is due to dry matter loss and on the other to water absorp- 

 tion. If expressed as a percentage of fresh weight, and the 

 percentage change is due solely to loss of dry matter, which 

 is the condition assumed by Chibnall, then with no real 

 change in nitrogen the loss of 1.15 g. of dry matter would 

 make a seeming change of nitrogen from 0.753 to 0.763 

 per cent, a seeming gain of 1.46 per cent over the orig- 

 inal nitrogen. If the change in the dry-matter content 

 expressed as percentage of fresh weight is partly due to 

 water absorption and partly to dry-matter loss, as is more 

 likely, the change in nitrogen content, which is apparent 

 and not real, would be between the extremes here given, 

 that is between 1.46 and 7.6 per cent. 



Although his figures show a high percentage error by 

 method A, and low percentage errors by methods B and C, 

 there would be no systematic error due to lapse of time 

 by method A, whereas changes in either dry weight or 

 fresh weight with time would introduce errors into both 

 methods B and C. 



Data obtained from carefully matched leaves, which 

 Denny has called the ''twin leaf" method, are much more 

 dependable for determining real diurnal variations. By 

 the use of this method Denny (1930) found no indication 

 of a significant diurnal variation in the nitrogen content 

 of Salvia leaves. Gouwentak (1929) by the half-leaf 

 method found no consistent diurnal variation in Helianthus. 

 Although the method is superior to others for determining 

 changes in absolute amounts of water, nitrogen, ash, etc., 

 in specific organs such as leaves and fruits, it is obvious 

 that the changes may not be strictly normal. Removing 



