31 



difficult!}' soluble, probabl}' organic. Opposed to the first explanation 

 is the fact that at no time while the soil was standing, or on the addi- 

 tion of the nitrate solution, was any evolution of gas ol)serve(], and as 

 0.86 gram of moist nitrogen measure 792.6 cubic centimeters at 30° C 

 and 760 millimeters, it is not conceivable that even a small portion of 

 this amount could escape as gas without detection. In support of the 

 second explanation is the nitrogen content of the soil. The original 

 soil contained nitrogen 0.987 per cent and moisture 16.65 per cent, 

 which is e(|uivalent to 1.184 per cent nitrogen in the moisture-free 

 soil. The figures obtained from the wet soil in the percolator after 

 treatment were nitrogen 0.64 per cent and moisture 49.30 per cent, 

 which is ecjuivalent to nitrogen in the moisture-free soil 1.262 per 

 cent. The difference obtained from analj^ses of the soil 0.78 per cent 

 nitrogen is reasonabh' near the difference between nitrogen added antl 

 nitrogen recovered, viz, 0.86 gram in 1,000 grams of soil or 0.()86 per 

 cent. 



There is, of course, nothing new in the recognition of the phenome- 

 non of denitrification. It has been generalh' recognized as taking 

 place in soils, especially those rich in organic matter, when for any 

 reason there was not free access of air, as in the presence of excess of 

 water. Various substances such as fresh manure, straw, etc., have 

 been stated as being important factors in the case. Several investiga- 

 tors have also pointed out that in some cases there are good grounds 

 for the assumption that in this change of nitrates a portion is stored 

 in the soil in a little soluble, probably organic form. 



Most writers on the subject make the statement that under ordinary 

 conditions of cultivation denitrification is not an important factor in 

 the growing of crops. While this may be true, generally, there are 

 certain phases of the situation in Hawaii, notably climatic conditions 

 and irrigation practices, which, in view of the preliminary work out- 

 lined above, force the conclusion that it is an important question here. 

 In most localities here where crops are grown depending on the rain- 

 fall alone, there are times when the soil is practically' saturated wath 

 water for weeks at a time. Where irrigation is practiced, limited for 

 the most part to cane growing, it has been shown by several workers 

 here that the soil is practically at or near the saturation point through- 

 out the irrigating season; conditions which, as shown, l)ring about the 

 denitrification of nitrate present or added to the soil. It has for a long 

 time been accepted as axiomatic, especially by the planters here, that 

 heavy rains or excessive amounts of irrigation water wash all nitrates 

 out of the soil unchanged. This the work outlined above has demon- 

 strated to be a mistake. For instance, in the experiment with 1,000 

 grams of soil noted above, the amount of nitrate added was equivalent 

 to 15 tons of nitrate of potash per acre, using the weight of soil to the 

 depth of 8 inches as the basis of calculation. But after the addition of 



