“…The analysis of soil and plant samples needs to be carefully planned, particularly in fertilizer work, and where analytical facilities are limited greater emphasis should perhaps be placed on the soil sample, for annual crops, and on the plant sample for perennial crops (Robinson, 1967c), though both have a part to play in an effective advisory service (Scaife, 1968(Scaife, , 1969Robinson, 1968a, 1969a, b andFoster, 1969). It is advisable to arrange for the collection of on-site rainfall data, to use in running moisture deficit calculations (Dagg, 1965;Dagg and Tapley, 1967;Dagg et al, 1968) since, on this criterion alone, trial sites may be classified into broad categories on the basis of moisture availability over critical periods of crop growth, thus enabling greater precision of calibration to be attained.…”