depth increments in centimeters will be: 0-2, 2-5, 5-10, 10-15, 15-25, 25-35 and at 15 cm intervals below 35 cm. If a soil horizon is encountered the interface lines will be chosen as the increments from the horizon rather than the fixed nominal increments from the surface. Locations for the collection of soil samples will be chosen on the basis of (1) random selection and (2) ground or aeri#l] survey meter readings. Samples collected on the basis a A. 1. oO These are described below: of randcon:n relect on, Each island will be divided by a narrow grid, approximately 2500 points or 50' x 50! smaller). 2. (i.e., grid whichever is All areas defined by such a grid vill be numbered, if stratification is desired and in orier, the strata will be chosen and indicated on the prid netyvork, 3. Individual semple areas will then te chesen within each stratification or grid by selection of the cpnmropri:te coordinate number utilizing rendom number usbles. The unmber of samples per island or group hes been previcusly determined through consideration of desired 4. goals and statisticil significance, The exact location of the sample collection is to be the center of the area chosen by the random number Lechniaue described in No, 3 above. ft is realized that the deternnition of such a point with any great precision or eccursey in the field is technically imprectical in most cases. It is mect important however that the sexple collecter mute ever; vrecsontbie effort to locate the position as closely as possible. nn perticular, the sample should come from within a 1O foot by 19 fect area defined as the center area of the grid point. Tdcally, the semple will come from the exact center of this Limited crea as just stated. In the field the location will be identificd as indicated on the man but will 4 be located prot cbly by pacing or other field direction. The spot so determined by such preinmg will to Uhe «cturl spot at the end of the decipnuete? meber of pacer, nnd ne otter. If there is some obstacle ts cuwmliny at thos pee Pied Jocation (e.c., 2 con- crete psd) then that Snot shosks ic recorde? in the field aud no sample taken,