MICRO-NUTRIENT PROBLEMS 33 



It is also possible that such a 'wet ashing' method may be 

 preferable to dry ashing when micro-nutrients other than copper 

 are to be estimated. In this connexion it may be mentioned that 

 Griggs, Johnstin and Elledge (1941) found that dry ashing at 

 400° C. resulted in a loss of 30 per cent of the total potassium 

 and 10 per cent of the total calcium. They themselves recommend 

 the extraction of the mineral elements with nitric acid and 

 perhydrol. Nitric acid is added to a weighed quantity of the 

 plant material in a 70 ml. pyrex test-tube and the mixture 

 heated on a sulphuric acid bath at a temperature between 120 

 and 140° C. until the solution is clear. Perhydrol is then added 

 and the tube carefully heated over a micro-burner. If necessary, 

 small additions of nitric acid and perhydrol may be made to 

 effect complete decoloration of the solution. 



The best established of the trace elements are, as we have seen, 

 manganese, zinc, copper and boron. The determination of each 

 of these in plant material will now be considered. 



Manganese. A considerable number of methods of reasonable 

 accuracy are available for the determination of manganese in 

 small quantities. The spectrograph, polarograph and absorptio- 

 meter can all be employed successfully for this purpose. With 

 regard to the spectrograph, the flame, arc and spark methods can 

 all be used, but, as has already been indicated, the sensitivity of 

 the flame method is less than that of the arc. For the deter- 

 mination of manganese in plant material the flame method has 

 been used and described by Lundegardh (1929, 1934) and more 

 recently by Griggs, Johnstin and Elledge (1941). As already 

 mentioned, by its means about 1-4/xg. of manganese can be 

 determined. Lundegardh claims that the probable error of a 

 single determination made by the flame method is about 1-2 per 

 cent, but the accuracy is no doubt less than this as the amounts 

 determined approach the lower limit of measurable quantities. 

 The line used for the measurement is 4031 A. (Lundegardh, 

 1929). Considerably smaller quantities of manganese can be 

 determined by the arc and spark, but, although they have been 

 used quite extensively in metallurgical work, they have received 

 relatively little attention from the point of view of the deter- 

 mination of manganese in plants. 



However, Melvin and O'Connor (1941) have used the arc 



