H. B. Brooks: The paper by Kinnard and Faus is important and valuable, not only for its definite data and clear reasoning, but also because it is a tangible indication of the interest which is being taken in the question of the accuracy of metering. This interest reveals itself in many ways; for example, in the increasing attention which is being given to the ratio and phase-angle performance of instrument transformers. Among reasons for this interest may be mentioned the greater attention which must be given to plant operation, because of higher costs of fuel and other essential supplies; the growing tendency toward interconnection, with the consequent need of accurately measuring interchanged energy; and the necessity for accurate electrical measurements in acceptance tests of turbo-generator units, where an error of a fraction of one per cent may mean an error of thousands of dollars in penalty or bonus. Such tests must necessarily extend over a period of hours, and the load cannot be kept as constant as might be done with small machines in the laboratory. It is only natural, therefore, that those having to make such tests should turn to the watthour meter as a means of measuring the energy and hence the electrical power output. However, from information which has come to us, it appears that such use of watthour meters does not give results as consistent as can be had by the use of indicating wattmeters read at sufficiently frequent intervals. Nevertheless, the large amount of labor required for the latter procedure makes it desirable to bring up the performance of the watthour meter, if possible, to such a stage that it will be good enough for plant acceptance tests. The work done by Kinnard and Faus is an important step in this direction, and I hope they will continue their analysis of sources of meter errors and means for overcoming them.