tive forces of nature, such as sunshine, rain-fall, frost, dew, heat, cold, light, darkness, etc., or to those agencies which are frequently present in atmospheric air, that shorten the life of dried paint, such as acid or alkaline gases and vapors. Comparative tests on limited surfaces-that is, small panels of steel, iron, or glass-may be misleading because the painter, unless experienced in making such tests, cannot gauge his work with unfamiliar material, and because his brush may not be in condition to give the paints an impartial trial.

However, given a properly equipped testing-room, a thoughtful painter trained by long and constant practice in making com- parative tests on small panels, and a sys- tematic method faithfully pursued of ex-amining and recording conditions and results; we  believe  tests  of  an  unfamiliar  paint  or