From time to time experiments have been published in regard to the quality of seeds. Some are bought from one place, some from another, and some from somewhere else. These are counted and sown side by side; fifty grow from one sample, sixty from another, seventy-five from another, and the results published to the advantage of one or the other as the case may be. But these single experiments often do great injustice. There are scores of reasons why seeds do not grow besides their being bad. The only fair test of seeds, is to take five or more different lots of the same seed, and take the average for comparison for competition with a similar set from the other kind.