8 i8 
Journal of Agricultural Research 
Vol. V, No. 17 
SUMMARY AND CONCLUSIONS 
For the sake of convenience, the more important conclusions regard¬ 
ing the method of making penetration tests, which have been reached 
as a result of this investigation, are summarized below. 
(1) Melted samples should be cooled for not less than 2 hours prior to 
test, and should be tested upon the same day that they are melted, 
preferably after 2 or 3 hours. 
(2) Samples should be maintained at ti^e testing temperature for not 
less than 1 hour, and preferably for hours prior to test. 
(3) Upon standing in the air, prepared samples show a decreasing 
penetration, but no definite end point or set is produced up to 28 days. 
(4) In ordinary laboratory work there is no apparent advantage in 
cooling samples in ice or ice water prior to determining their penetra¬ 
tion at higher temperatures. Cooling in ice water is therefore not 
recommended. 
x 
(5) Samples should be maintained and tested within o.i° C. of the 
desired temperature for accurate work, as a variation in temperature of 
less than 0.5 0 in temperature may produce a decided difference in 
results. ' 
(6) Tests at 4 0 are not the practical equivalent of properly made 
tests at o°. 
(7) When making tests at o°, samples should not be packed in 
crushed ice, but should be immersed in a brine bath. 
(8) The increase in penetration of a material determined under given 
conditions of temperature and time is, within certain limits, almost pro¬ 
portional to the increase in load. For the 100- and 200-gm. loads varia¬ 
tions of as much as 1 gm. do not as a rule seriously affect determinations. 
It is, however, recommended that in all cases the load should not vary 
more than 0.2 gm. from that desired. 
(9) In any test, proportionally the greatest number of points penetra¬ 
tion is obtained during the first second. In the 5-second test approxi¬ 
mately one-half of the total penetration is obtained during the first 
second. A variation of one-half second may, however, produce an appre¬ 
ciable variation in results. 
(10) A carefully calibrated metronome is recommended for securing 
the proper time control. 
(n) Aside from possible variations in needles, it is believed that va¬ 
riations in results obtained upon the same material by different labora¬ 
tories are more probably due to unobserved variations in the methods of 
preparing the sample and to the control of temperature than to any 
other causes. 
(12) It is believed that a study of the penetration of various types and 
grades of bituminous materials under a variety of conditions of tempera¬ 
ture, load, and time may throw considerable light upon their other 
physical and chemical characteristics, and may serve as a possible 
means of identifying their origin and method of manufacture. The 
writers propose to continue work along this line. 
