Abstract :
W. H. Pratt: I think it should be emphasized that the present paper, except so far as it describes a particular mechanism for effecting compensations, is a restatement of material most effectively presented by Kinnard and Faus at the midwinter convention two years ago. Previous to that time, meters having much smaller Group II errors than those used for a background of this paper were used in predominant numbers, and since then the whole output of one of the largest producers of meters has been substantially without Group I errors. So that, presented at this time, the background of this paper must not be looked upon as a picture of general present-day practise.
