_Government Reforestation Policy._ The duty of the Forest Service to put the denuded areas which will not be reforested naturally into a condition of productivity admits of no further argument. But the problem is not so easily solved as it is made clear. Under the semi-arid conditions prevailing on many National Forests this work involves uncertainties and unsolved problems. On the National Forests artificial reforestation was an untried field when the Forest Service entered it. The Government therefore had to develop its own practice in the face of a great variety of conditions, largely unfavorable. The situation still calls for intensive experiments to develop the best methods from the standpoint of both cost and results. More than that, it calls for a different set of methods for each forest region of the West which has its peculiar trees, climate, and soils. Then, lastly, when the proper methods have been demonstrated by experiment, the new methods can be applied on a large scale with a very good chance for success. Therefore intensive experiments must come first. Business prudence requires the development of all methods in detail and reasonable certainty as to their results before large sums are expended upon field operations. In the least favorable regions like the semi-arid mesas of the Southwest, the work is restricted for the present to small, carefully conducted experiments, the result sought being reliable information upon how to proceed rather than the reforestation of many acres. In the most favorable regions, as the western slopes of the Rocky Mountains and the Cascade Ranges, the results already obtained have been so excellent, due to an unusual combination of good growing conditions, that operations upon a larger scale have been justified simultaneously with continued intensive investigations. As the work is extended into each new region or new National Forest, the most favorable sites are always chosen first. After the possibilities and limitations of each method have been ascertained by experience under the best conditions of each locality the work can either be intelligently extended or restricted. But the work is always conducted from the standpoint of the maximum return for each dollar expended.