SOILS OF THE ARID AND HUMID REGIONS. 399 



of the soil's intrinsic stores of plant-food, that the need of 

 artificial fertilization is there largely restricted to those soils 

 of which the parent rocks were exceptionally deficient in the 

 mineral ingredients of special importance to plants, that or- 

 dinarily form the essential material of fertilizers. Quartzose, 

 magnesian, and other soils resulting from the decomposition 

 of " simple " rocks will, of necessity, be poor in plant-food 

 everywhere. 



Humus in Tropical soils. Another inference from the cli- 

 matic conditions of the tropics is that the properly tropical 

 soils are likely to be rich in humus, as a result of the luxuriant 

 vegetation which in the decay of its remnants must leave abun- 

 dant humic residues. This seems to be generally verified 

 wherever the interval between rainy seasons is not too long; 

 for otherwise, under the great and constant heat of the tropics 

 a rapid burning-out of the humus, such as is known to occur 

 in the arid regions, must also take place. A good example 

 illustrating the intertropical regime as regards humus is given 

 in the table in chapt. 8, p. 137, showing the humus-content of 

 some Hawaiian soils. Both are of the same order as in the 

 soils of the temperate humid region, though the nitrogen-con- 

 tent evidently can, consistently with productiveness, range 

 lower than has thus far been observed in temperate climates. 

 This again forms a striking contrast with the soils of the arid 

 regions. 



It is greatly to be regretted that not even approximate de- 

 terminations of the organic matter, much less of the humus- 

 substance proper, have been made by any of those who have 

 analyzed tropical soils; excepting those made of Hawaiian 

 soils at the California Experiment Station. 



The " loss by ignition " is of courst- always very largely 

 water, mostly referrible to ferric hydrate and clay substance, 

 the latter presumably essentially in the form of kaolinite. 

 \Yhen, therefore, ferric oxid and alumina have been deter- 

 mined, we may approximate to the amount of total organic 

 matter by making allowance for ferric hydrate at the rate of 

 about 14% of the ferric oxid. for kaolinite at that of 34.^2' o 

 of the alumina found. Deducting these amounts of water 

 from the total " loss by ignition," we may obtain at least an 

 approximate idea of the organic matter, and the probable 



