302 cox. 



tlie constituents of the atmosphere contribute to plant food may be, the 

 total composition of a soil must always ultimately be taken into account, 

 for from this under normal conditions certain elements of plant- food are 

 derived. Certain constituents become exhausted after years of cultivation 

 and the soil requires fertilization in order to make it productive. This 

 is especially true of the elements nitrogen and phosphorus. When the 

 percentage of any of the elements of the plant-food existing in the soil 

 falls below 0.1 the productive capacity of the soil may be questioned. 



Such chemical and physical characteristics of Philippine soils as I 

 have been able to segregate are given below. Although incomplete, these 

 data should aid in the selection of suitable soil for certain crops, in the 

 reduction of the number of individual failures, and in the future as the 

 beginning of a basis for the study of crop yields. It is hoped that they 

 will be of assistance to teachers in connection with their school gardens. 

 Analyses made in the Bureau of Science of Philippine soils from many 

 different provinces are given and as much additional information as the 

 author has been able to secure. It must always be borne in miud that no 

 analysis is more accurate than the sample which it represents so that the 

 errors of sampling should be reduced to a minimum. I have prepared as 

 an appendix to this paper directions for taking soil samples which are 

 satisfactory and for the sake of uniformity they should l^e carefully 

 followed. 



CHEillCAL ANALYSES. 



All analyses were made on that portion which passed a 1-millimeter screen 

 ("fine earth"). In the ease of Philippine soils, the detritus on a 1-millimeter 

 sieve is usually very small and I believe the portion passing a sieve of this size 

 contains practically all the constituents from which the plant derives its 

 food and includes all that should be termed "fine earth" or soil.' The methods 

 used were substantially those of the Association of Ofiieial Agricultural Che- 

 mists.^" With the exception of moisture the results are given on the basis of 

 a sample dried to a constant weight at 105°. In most cases the moisture was 

 determined on the air-dried sample." The results are as follows: 



° An inspection of the results of G. M. MacNider, Journ. Ind. Eng. Chem. 

 (1909), 1, 447, shows that in nearly all cases the potash and in many cases the 

 lime and nitrogen are entirely contained in the soil passing an 0.5-niillimeter 

 sieve, but that the phosphoric anhydride is more uniformily distributed. If 

 there were to be a large detritus on a 1-millimeter sieve the results with respect 

 to potash and in many cases to lime and nitrogen might appear high and would 

 not fairly represent the composition of the soil. 



" U. S. Dept. Agri., Bur. Chem. Bull. No. 107 (Revised). 



" The -present practice of the laboratory is to use the dry weight as the basis 

 oh which to compute the percentage of moisture also. The advantage of comput- 

 ing percentages on this basis is that it furnishes a constant means of comparison, 

 whereas if computed on the actual or wet weight the basis would vary with every 

 change in the amount of moisture. 



