Curie–Weiss from the spin count
The Curie constant is fixed by the number of unpaired spins n per formula unit; the formula weight converts the molar result to a per-gram susceptibility. Everything else — the Weiss temperature and the temperature-independent term — is yours to drag.
χ(T ) =
C T − Θ+ χ₀
C =N A g ² S (S +1) μB ²3k B , S = n /2
χ vs T
Susceptibility, 2–400 K
1/χ vs T
Inverse susceptibility, 2–400 K
High-T limit of 1/χ
—
1/χ → 1/χ₀ as T → ∞
C = N A g ²S (S +1)μB ²/3k B = 0.12505 μeff ² cm³ K mol⁻¹ with μeff in μB . For g = 2 this collapses to the familiar C = n (n +2)/8, i.e. 0.375 for one unpaired spin and 4.375 for five. n is the count per formula unit , so a compound with two S = 5/2 ions per formula unit needs n = 10, not 5.
The formula weight changes no physics: it only converts the molar susceptibility to a per-gram one, χg = χmol /M . Switching the toggle rescales both panels (and χ₀ with them); Θ and the shape of the curves are untouched.
With χ₀ ≠ 0 the inverse susceptibility is not a straight line: it curves away from the Curie–Weiss line and saturates at 1/χ₀. Fitting 1/χ with a straight line over a finite window therefore returns an apparent Θ and C that both drift with the fitting range — the reason a χ₀ term is usually carried explicitly in the fit. A negative χ₀ drives χ through zero at T * = Θ − C /χ₀, where 1/χ diverges; that pole is drawn as a break in the curve. Spin-only C is itself an assumption: unquenched orbital angular momentum, crystal-field effects and covalency all move μeff off √(n (n +2)), which is what the g slider is for.