HO DEPARTMENT BULLETIN 1216, U. S. DEPT. OF AGRICULTURE 
direct method of verifying testing machines shall apply to the verification of 
testing machines by the comparison method so far as is practicable. 
Note. — The method of verification of a testing machine by standard weights (section 4) 
and the method of verification by standardized proving levers (sections 5 to 8) are both 
regarded as superior to the method of verification by the use of companion specimens. If 
the results of a verification test of a testing machine by the comparison method fail to 
agree with the results of a verification by either the standard-weight method or the 
proving-lever method, the results of the verification by the comparison method shall be 
discarded. 
18. (a) A testing machine which is to be used as the standard for verifying 
other testing machines shall itself be verified by a special test, in which the veri- 
fication is carried out twice. The loading range shall be determined by the 
tolerance for new testing machines, namely, ±1 per cent (section 22). For the 
loading range established, the differences in indications found in the two tests 
of the standard machine shall not exceed 0.25 per cent. 
(&) In verifying testing machines by the comparison method, corrections shall 
be applied to the indications of the standard machine and these corrections shall 
be based on the average correction obtained in the two tests used in its veri- 
fication. 
19. (a) For each test load, eight or more tension test specimens shall be cut 
from soft rolled or drawn steel and numbered consecutively, as in Figure 
47, (a). 
fa) [T 
E 
16 17 
I 
,- Radius not less than 
(b) 
Rough Turn 
i^_ 
,d 
Rough turn 
\\ , II 
8d ->j \<-4+d->\ K- 
8d 
Fig. 47. — Comparison specimen: (a» Method of numbering specimens; (6) form 
of test specimens 
(&) The tensile strength of the steel shall be determined by a preliminary 
test and the sets of specimens for the comparison test shall have such nominal 
cross-sectional areas as will give, approximately, the loads required. The form 
of the specimen shall be as shown in Figure 47, (&). The ends may be threaded, 
or otherwise machined, to fit holders, but it is recommended that those speci- 
mens shipped to another laboratory be left with cylindrical ends. The actual 
value of "$" (fig. 47) for each specimen shall be determined by means of a 
micrometer. 
Note. — It should be noted that the standard tension test specimen, 0.5 inch in diameter 
and having a 2-inch gauge length, satisfies the requirements for a specimen for the veri- 
fication of a testing machine by the comparison method. See 1924 Book of A.S.T.M. 
Standards (I, p. &, fig. l). 
20. (a) The specimens having odd numbers shall be tested in the machine to 
be verified and those having even numbers in the standard testing machine. 
The tensile strength only shall be determined. 
(b) The speed of the testing machine, by which is meant the speed of the 
pulling head when the machine is running idle with no specimen in the machine, 
when testing verification specimens shall be approximately the same for both 
the standard testing machine and the machine being verified. The speed shall 
be the slowest speed that can be obtained on both machines; in any case the 
speed shall not exceed 1 inch per minute per inch of diameter of comparison 
specimen. 
(cy When testing verification specimens, especial care shall be taken to main- 
tain the weighing beam of the testing machine in equilibrium as the tensile 
