6.9Fitting Exponential Models to Data

The figure caption explicitly classifies one curve as ‘exponentially decaying’ (first order, quartz surface) and the other as ‘linear’ (zero order, tungsten surface) from the plot shape alone, before any fitting is done — exactly the scatter-diagram shape-recognition (linear vs. exponential vs. logarithmic) that is §6.9's first objective, not the choose-a-transformation move already bridged to §4.4.