8 i6 
Journal of Agricultural Research 
Vol. V, No. 17 
Upon reviewing these results it will be noted that a variation of i gm. 
in no case produces an appreciable variation in results. In fact, the 
greatest variation is found to be one point penetration, and, in many 
cases, no difference in penetration is to be observed. It is therefore 
obvious that errors due to the calibration of the weights are practically 
negligible. 
In connection with the series of tests for any individual material, it is 
of interest to note that within certain limits the increase in penetration is 
almost proportional to the increase in load. In other words, practically 
a straight-line curve may be obtained by plotting for any material the load 
against the corresponding penetration and connecting these points. If 
this is done the projection of the line to the axis representing increments 
of load will not hit this axis at its intersection with the axis representing 
increments of penetration. In general, it appears that blown asphalts 
possess less surface tension and adhesiveness than steam-distilled 
asphalts. The penetration of a blown asphalt therefore represents more 
nearly the actual distance which the needle enters the sample. In the 
case of steam-distilled asphalts the surface of the sample is markedly 
depressed by the needle, and probably proportionally greater retardation 
of its movement is produced by material which adheres to it. 
It is of interest to note that a steam-distilled asphalt having a higher 
penetration than a blown asphalt at 25 0 C. under a load of 50 gm. applied 
for 5 seconds may have a lower penetration than the same blown asphalt 
at 25 0 under a load of 100 gm. applied for 5 seconds. For this reason the 
relative penetrations of different types of asphalt do not necessarily 
indicate their relative hardness. 
As would naturally be supposed, in general, the greatest variations 
in penetrations due to variations in load are obtained upon the softer 
materials or those showing the highest penetration at any given load. 
The blown products, however, show more variation than do the other 
types. This is probably due to the fact that the effect of surface tension 
and adhesion is less pronounced with the blown products than with the 
steam-distilled products. 
It was thought unnecessary to study the effect of variations in load at 
other temperatures and for other periods of time, as there was no reason to 
suppose that the results would be different in character from those given. 
The changes in time and temperature would merely change the pene¬ 
tration of the material and should give results comparable with those 
obtained upon softer or harder grades of the same type. 
EFFECT OF VARIATIONS IN TIME 
Penetration determinations are ordinarily made for a period of 5 sec¬ 
onds, especially where the 100-gm. load is employed. In the case of 
materials which are quite hard they may be made for a period of 1 min- 
