in any particular case, it will be obvious that the power of an A.C. must depend on its e.m.f. and strength. But how shall we define either, since both are continuously varying? Shall we, in defining the strength of an A.C., give its crest value as so many amperes, or shall we give some smaller value? Apparently the simplest plan would be to give the crest value, but this would not be a true measure. If we take by way of illustration the case of a lamp lighted by passing an A.C, through the filament, we have seen that twice during each cycle there is a moment when the current is zero. At those times the lamp receives no power, whilst at the times when the current has crest value it receives a maximum of power. The average power absorbed by the lamp must therefore be something between this maximum power and zero. If, then, we define the strength of the current by stating its crest value, we overestimate it. The proper basis for estimating the strength of an A.C. is obviously that of equal effect produced by a D.C., and we may thus speak of the "effective" (sometimes also called the "virtual") value of an A.C. With modern A.C. machines the shape of the e.m.f. and current curve shown in Fig. 19 closely approaches a
