Barus and Strouhal— Viscosity of Steel. 25 
Discussion. 
Viscosity and temper.—l. The results of tables | to 26 may 
be discussed in two ways: We may either accept some definite 
and applicable law like that of Weber or of Koiilrausch, and 
calculate the mean constants for each set of results; or we may 
construct them graphicaily and then calculate the codrdinates 
of the mean curve for each set. The latter is decidedly the 
better way, because it is less arbitrary and more convenient. 
In table 83 the mean results in question are thus summarized. 
The first three columns indicate the tables from which in each 
case the selections are made, the rods selected, and their tem- 
pers. The remaining columns contain the differences of viscous 
detorsion, ¢,, in radians per centimeter of length of the bifilar, 
one of the wires of which is invariably glass-hard, the other 
annealed as stated. 4, is arbitrarily fixed at zero, for one hour 
after a twist of +180° and —180° has been imparted to the soft 
and hard wire, respectively. 
TABLE 33.—Digest of the mean results, Tables 1 to 26. 
oo x 108 in radiaus per centimeter of 
Table No. Rods Nos. (aes Haan en 
50" | 1004|150% 200/250" )300" |350" | 400 
1,2 1,2 20°, oP | 0-08] 0-11] 0-15 0-19]..2.J-.-_|---_|_-.- 
47, 43 (glass-h’d) | 
3, 4,5, 6,7, 8 | 
BA, 6A, 7,8 | 9, 10, 43, 44,| 100°, 10» | 0-59) 0-96] 1-41) 1°58) 1-71/_-. |... |_- 
45, 46 
| | 
| 
10, 13, 13A, 14 ae eee bol 190°, 18 | 1-38] 1-89] 2-27] 2-63] 2-95] |... |_--- 
? ? ? 
16, 18 yu o5 | 360°, 1% | 1-80] 2-61] 3-13] 3-53] 3-84] 4-14] 4-49| 4-70 
2 
| 9) 
19, 20 | je’ 5g | 400°, | 2°01| 2°76) 3-31) 3°75] 4-14| 4-49] 4-80] 5-07 
25, 2 | 
27, 28, 29 ; | 
23, 24, 25,26 |1 30, 31,32 | 1000° | 1-76] 2°51] 3:10|3-61) 4-04) 4-40] 4-67]... 
33, 34 eae 
All radii identically 20==0°082°™. Like signs of ¢ and of w refer to the same 
angular direction. 
The results of table 33 are graphically constructed in figure 
3, time in hours as abscissa, difference of angular detorsion g,, 
in radians, as ordinate. 
Table 23 and figure 8 lead to this curiously remarkable re- 
sult: If we abstract for the moment from the states of temper 
extreme hard and extreme soft, the viscosity of steel decreases in 
proportion as the hardness of the metal increases. Wecall to mind 
