SOILS OF THE ARID AND HUMID REGIONS. 383 



leaves and stems in the form of carbonate. But others seem 

 to possess this power to a limited extent only. 



In the case of soils containing much magnesia the proper 

 proportion between it and lime may easily be disturbed by the 

 greater ease with which lime carbonate is carried away by car- 

 bonated water into the subsoil, thus leaving the magnesia in 

 undesirable excess in the surface soil. Hence the great ad- 

 vantage of having in a soil, from the outset, an ample propor- 

 tion of lime. From this point of view alone, then, the analyti- 

 cal determination of lime and magnesia in soils is of high 

 practical value. 



Aso, Furuta and Katayama (Bull. Coll. Agr. Tokyo, Vol. 4 No. 5 ; 

 Ibid. Vol. 6), have by direct experiment determined the most advan- 

 tageous ratio of lime to magnesia in several crop plants. They find 

 for rice and oats i : i, for cabbage 2 : i, for buckwheat 3:1; there being 

 apparently a connection between the extent of leaf-surface and lime 

 requirement, since leaves contain predominantly lime, while in the fruit, 

 magnesia predominates. 



Manganese. A decided difference in the manganese content 

 of the arid as against the humid soils appears in the table, the 

 ratio being about n : 13 in favor of the humid soils. Man- 

 ganese has not been regarded as being of special importance to 

 plant growth in general, although, as already stated, some 

 plants contain a relatively large proportion of manganese in 

 their ashes ; thus, e. g., the leaves of the long-leaved pine of the 

 cotton states. 1 But no definite data showing the importance of 

 this element to crops were available until Loew and his co- 

 workers at Tokyo 2 established its stimulating action in a num- 

 ber of cases, in which crop production was materially increased 

 by the use of protoxid salts of manganese. Aso 3 applied man- 

 ganous chlorid to an experimental plot of thirty square meters, 

 at the rate of twenty-five kilos of Mn -3 O 4 per acre, and thus 

 obtained a yield of rice one-third greater than on the control 

 plot, at a cost of about $2.00, while the value of the increase 

 of the product was nearly $68.00. More experimental evi- 

 dence on this subject is required to establish the general value 



1 Rep. Agr. and Geology of Mississippi, 1860, p. 360. 



* Bull. Agr. Coll. Tokyo, Vol. V., Nos. 2 and 4. 8 Ibid. Vol. 6. 



