34 Barus and Strouhal— Viscosity of Steel. 
of a unique maximum of magnetization seems certainly to be 
coincident with the occurrence of maximum viscosity. The 
magnetic maximum so far as our results go is apparently much 
more clearly pronounced than the viscous maximum. We 
remark in general that as the ratio of length to diameter 
increases, the minimum of permanent magnetization shows a 
tendency to move from soft to hard. Furthermore, it is proved 
that the permanent magnetization of soft rods is greater in pro- 
portion as they are more and more vearly linear.* Hence as 
our march approaches the linear limit we observe an unmis- 
takable tendency toward more detailed similarity between the 
variations of the magnetic and the viscous properties of steel. 
Mr. Hopkinsont in his memoir on the magnetization of iron 
introduces a definition of coercive force, defining it as “that 
reversed magnetic force which just suffices to reduce the 
induction to nothing after the material has been subjected 
temporarily to very great magnetizing force.” There is a slight 
objection to this definition, inasmuch as it introduces a some- 
what vaguely complex state of zero-magnetization.t . But the 
feature in question is as hard to improve as the definition is 
valuable. Mr. Hopkinson’s extensive experiments show that 
coercive force is invariably increased by hardness. Together 
with others we Lave proved that hard steel is alike well 
es to withstand the influence of percussion. 
These general relations between viscosity and maximum 
permanent linear intensity observed for steel are sustained by 
iron. According to data “contained in a paper of the lamented 
Dr. L. M. Cheesman§ tie permanent magnetization of drawn 
iron exceeds that of soft iron. The magnetic intensity of per- 
manently saturated drawn iron is comparable with that of soft 
steel and quite equal to that of hard steel. More recently 
and even more elaborately Hwing has discussed the subject. 
He finds ‘‘in the absence of mechanical or other disturbance, 
soft iron is far more retentive than either hard iron or steel,” 
an exceedingly remarkable result.|| Our experiments show 
that the intensity of applied stress is an important factor in 
determining the viscous behavior of iron; that for moderate 
* Bull. 14, pp. 240 to 143. 
+ Hopkinson: Phil. Trans., ii, 463, 1885. 
{ The unmagnetic state, though dependent on the details of the process of tem- 
pering, is a distinct magnetic state. Similarly we may regard the saturated mag- 
netic state, though allowance must here be made for the occurrence of cyclic 
magnetization, a phenomenon suggested by Fromme (Wied. Ann., iv, p. 89, 1878) 
and (Wied. Ann., xiv, p. 308, 1881), concisely interpreted by Warburg (Wied. 
Ann., xiii, p. 141, 1881) and which Ewing (Phil. Trans., ii, p. 545, 1885) has re- 
cently discussed generically under the name of “ hysteresis.” 
§ Cheesman: This Journal, xxiv, p. 183, 1882. Lengths 100 diameter: and 
less. 
|| Ewing: l.c., p. 541. 
