considered to be uncertain, or where the gap between points was so great as to cause difficulty in perceiving the continuity of the curve. From the original graphs the curves were copied over a light-table onto tracing paper. A somewhat natural sequence of subject matter was arranged, progressing from soil kingdoms. through the plant and animal Each graph was oriented by means of the 1000-day verti- cal line. Thus, the position of the curve in the series is no index of the amount of radioactivity involved. Absolute levels may be computed by referring to the column of Table 1 headed ''Plate count, 1000 days." For example, if for curve 25, the value of the plate count at 2300 days were required, then reference to Table 1 would show a plate count at Peet) 1000 days of 700, while the ordinates as scaled from Figure 1 on days 1000 and 2300 were 18 and 2.4 respectively. The ratio, 2.4/18, is multiplied by 700 to give a result of about 93 c/m, which is reasonably close to the actual value, 90 c/m (original data, not shown here). Slopes of curves were scaled from the finished drawings of Figures 1-6, using a transparent slope-scaler calibrated in tenths from slope -0.1 to -2.0 and by greater increments from -2.0 to -5. For each curve, the portions of markedly different slope were measured separately, thus necessitating subjective interpretations as to what constituted a change of slope.