The Legislature has directed the Public Service Commission to prescribe rules and regulations for the testing of electric meters (Public Service Law, § 67; Cons. Laws, ch. 48). In conformity with this mandate it has for many years required meters to be tested without removal from their positions of instalment. An electric meter is obviously a delicate instrument, and so the proof shows. The requirement of testing the same in position has proven over a period of years to give a maximum of accuracy in results. To grant the application of petitioner would require a change in the rules so as to permit the removal of a meter from its fixed position and the insertion of a so-called jack before testing. The Commission has found that such a method does not preclude the possibility of error. Quite evidently it has balanced a practice, known from actual use to be accurate, against experimental tests conducted largely by interested parties under conditions not clearly those to be anticipated in actual use. Moreover there were admissions by experts that the removal of meters might affect the accuracy of tests. The Commission had a right to require more convincing proof before abandoning a practice of known success.
Article 78 has its limitations. It should not be applied rigidly to proceedings of the character involved here. No burden was cast on the Commission to defend a regulation conceded to be successful, rather it is the quality of proof produced by petitioner as to the accuracy of a new method that is under scrutiny. The issue is whether*such proof is so overwhelming that its *332rejection must be considered arbitrary. The testimony as a whole leaves some room for doubt, at least, as to the efficacy of the new method. Under these circumstances the judgment of the Commission should not be disturbed, and our own views substituted in place thereof.