Guillaume compared alloys rather than assuming one steel
Guillaume’s systematic nickel–steel studies in the 1890s found compositions with exceptionally small thermal expansion. Invar made measuring rods and other precision parts less sensitive to ordinary temperature changes.
The useful result was a measured material anomaly. It depended on composition, treatment and the chosen temperature range, not on a universal claim that metal could stop expanding.
Sources: [1]
A small coefficient still accumulates over length
For a modest temperature interval, fractional length change is approximated by a linear expansion coefficient times temperature change. Long baselines turn small coefficients into measurable shifts. Temperature gradients can also bend or distort a part even when average expansion is small.
Sources: [1]
Elinvar addressed a different problem
A clock spring’s behavior depends on elastic response as well as dimensions. Guillaume’s later Elinvar work reduced temperature dependence of elasticity. Treating Invar and Elinvar as synonyms would confuse the length error with the mechanical timekeeping error.
Sources: [1]
The measurement context is reused
The 1920 award recognized the precision-measurement value of nickel–steel anomalies. The existing Michelson instrument explains how a small optical-path change becomes a fringe shift. It is linked as metrology context, with no claim that a temperature-driven Invar extension has already been implemented.