L. J. Wild 171 



In the case of the two Southland areas the explanation sought would 

 appear to be simply in differences in the calcium oxide contents of the 

 soils. The soils requiring Ume have an average of only 0-43 % of CaO, 

 while those not requiring Kme have 0-94 % on the average. But when 

 we come to the Canterbury soils, this explanation is insufl&cient, for, 

 among ten soils analysed by the writer, having a less acidity even 

 than the Southland soils not requiring lime, the average percentage of 

 CaO is only 0-55. The average lime content of nine Canterbury soils 

 analysed by Gray is 0-38 % ; of eight soils of this district analysed by 

 Paulsen 0-56 %, while 32 determinations by Aston yielded an average 

 of 0-50 % . The variations in the average lime requirements of the soil 

 groups shown in Table IX appear therefore to be no more correlated 

 Mdth differences in their Ume contents than do the practical results 

 of liming in the various areas. 



2. Calcium carbonate . Determinations of calcium carbonate have 

 not been made. None of these soils show marked effervescence with 

 dilute acids, and there is no external evidence indicating a connection 

 between lime requirement and calcium carbonate content. 



3. Lime magnesia ratio. A good deal of attention has been paid 

 in late years to the necessity for a proper ratio of hme to magnesia 

 in soils^. In this country Aston^ has recorded definite instances of 

 infertihty due to the toxic action of excess of magnesia, and has advo- 

 cated the use of Hme in certain soil areas, amongst them being some 

 Canterbury soils, for the sake of increasing the CaO : MgO ratio. His 

 own analytical results for the Canterbury Plains soils as shown in 

 Table X, Group F, scarcely bear out his contention in the case of these 

 soils at least; nor do the results of Paulsen's analyses of soils of the 

 Lincoln district, though Gray's analyses suggest a " bad " ratio. British 

 authorities find little evidence for the theory, Russell^ recording soils, 

 both rich and poor, having practically identical ratios. 



4. Potash. Nothing is revealed by the determinations of phos- 

 phoric acid and potash, for while Canterbury soils are slightly richer in 

 the former and markedly richer in the latter constituent than Southland 

 soils, yet the differences between the soils of the two Southland areas 

 are small enough to be considered negligible. 



5. Hygroscopic moisture : loss on ignition and soluble humus. An 

 examination of Table X will show marked and regular differences 



1 For a summary see The Plant World, Apl. 191(5, p. 83, et seq. 



- Aston, B. C, A'^. Z. Journal of Agriculture, Vol. xi, No. 6. 



* Russell, E. J., Soil Conditions and Plant Groivfh, 2nd ed. (1915). p. 1(35. 



Jouin. of Agric. Sci. viii i^ 



