We present precise experimental data suggesting that domains cooperate in f
erromagnetic hysteresis. This model has potential applications in magnetost
rictive transducers, transformer design, magnetoelastic effects and hystere
sis in general. For an annealed 3% nickel steel alloy, we measured magnetic
induction from the demagnetized state to near saturation, the reversal to
near saturation and return branches from both these curves. Modeling variab
les include the coercive field, saturation magnetization, initial susceptib
ility and the hysteresis effective field. Magnetization is scaled to its sa
turation value and magnetic field to its coercive value in the direction of
field change, and susceptibility is scaled to its reversible function. Sca
led differential susceptibility increases from its reversible value exponen
tially with magnetization change from any reversal and decreases exponentia
lly with decreasing slope to its reversible value at saturation. We define
the slope of this exponential to be a hysteresis constant divided by the sc
aled effective field, and this dimensionless field is the square root of 2
for hysteresis loops of polycrystalline domains with fully nucleated walls.
Rayleigh's hysteresis constant, valid for heavily worked steels over a sma
ll region, transforms to our cooperative hysteresis constant at larger fiel
ds. For the saturate curve, the scaled effective field reaches a minimum le
ss than 1 near the coercive field as domain interactions increase through w
all nucleation and their local fields. Variation in the effective field cau
ses hysteresis loops to creep upon cycling.