ABSTRACT
SUMMARY.-l'he method oj representing economic phenomWlu£ by curves demands close1' examination than we have yet given it, and tU1'1IS out on inspection to present many problems both oj interpretation and construction. Tlte measurements on the axis oj Y indicate limiting rates oj marginal significance, and, while expressed in a·n objective rate-unit, they 1nust ultimately rest on estimates based on psychic experience. Hence difficulties arise as to the relation between objective and psychic units, tke possibility oj keeping that relation stable, the meaning u'e are to attach to accuracy oj estimate and the conditions which limit that accuracy. If we express tke datn oj Book I. Ohapte1' lI. as to the significance oj tea in the jorm oj a ten curt'e 10e are led to 'exam-ine (a) t/u implicatiuns of tlte special formula to -which ou.r data. conformed, and (b) tlte possibility of any simple mathematical formtda approximately representing tlte facts. An attempt accu1'ately to inte1'pret tlte curve fU1-tlter leads us to distinguish between II cm've of total satisfact-ion and marginal s-ignijicllltCe 0'11 the one hand, amI a curve of pl'ice-and-qualttity-pu1'clwsed on the other hand. We find that these turvrs can, at best, only coincide approx'imately, (md tlutt all -indil'idual cur'l.·e purporting to represent both series of phenomena can theoretically only be a " temperamental" compromise.
