Since the energy partition in the two types of explosions is significantly different, particularly in the roughly 15 per cent of the Yield of a nuclear explosion which takes the form of prompt radiation, it seems necessary to consider an efficiency factor less than 1 for nuclear explosions as far as the cratering effects are concerned.* Experimentally, evidence on this point is meager in the extreme, being limited to the JANGLE surface and JANGLE underground shots. At this point it is useful to consider the numerical data on the JANGLE surface and the JANGLE underground shots. The data from these two shots can be placed on this curve with efficiency as a parameter; thus the curve DE on Fig. 4.3, represents the JANGLE surface shot for a radiochemical yield of 1.2 KT times the efficiercies shown on the curve, with radius scaled on the basis m = 3.4 and charge depth (height) scaled on the basis m= 3. Similarly the curve FG represents the JANGLE underground shot data on the basis 1.2 KT times the efficiencies shown there, using the same procedure. It will be seen that curve DE for the JANGLE surface shot intersects curve AB at an efficiency of about 60 per cent and that curve FG representing the JANGLE underground shot intersects the TNT curve at an efficiency of 107 per cent. It is not suggested that these values of efficiency are correct, but their comparative values are at least in the direction expected. It is recognized that, in accordance with the definition of+the equivalent TNT charge, the efficiency of the JANGLE surface shot should be defined as the value at the intersection of curve DE with the solid curve. It is nevertheless believed that there are such gross differences in mechanism between nuclear and TNT explosions in this region of close above-surface shots that the equivalence should be divided into two parts, one of which is concerned with the disparity in the form of the blast wave and the other is concerned with the remaining elements of efficiency. It is felt that the value of 107 per cent obtained on this curve for the JANGLE underground shot is probably unrealistic for tne following reason. It is clear that values of the scaling exponent m, and values of efficiency, can be paired to fit any crater measurement from a specific yield and depth. Since it is felt that efficiencies at greater depths than 17 ft should probably be nigher than at that depth and since it is also felt unlikely that nuclear efficiencies are nigher tnan 100 per cent, it appears that this value of efficiency for tne JANGLE underground snot is on the high * For present purposes, efficiency may be defined as the ratio of the total energy release of an equivalent TNT charge with that of a nuclear explosive. The equivalent TNT charge may be defined as the charge which at the same actual (not scaled) depth produces the same crater. Since in both TNT and nuclear explosions it seems reasonably established that only a small fraction of the total energy released can be accounted for in crater production, there is no pnilosophical reason why the efficiency of a nuclear explosion as defined above need be limited to 100 per cent; however, at all times of interest in the formation of craters the pressure within a nuclear explosion is higher than that within the equivalent TNI explosion and hence at the time venting takes place a greater fraction of the energy in a nuclear explosion should be dissipated to the air.