The purpose of this experiment is to test the relative yields of grain from across-bred grains s.s compared with self fertilized seed of the same varieties. It was originally purported to make these tests with both wheat and barley. However, owing to unfavorable greenhouse conditions the wheat plants bloomed very poorly and the crosses made set very little seed so the wheat was dropped and the work done entirely with barley. Before entering into the details of the work conducted, a brief review will be made of small grain breeding work at other places. The Agricultural Experiment Station of Wisconsin has only recently issued a bulletin on barley cultivation. Barley being one of the principal crops of Wisconsin, the Experiment Station has made considerable effort to improve the seed production. Individual selection and crossbreeding have been carried on extensively. Up to the present time no cross-bred varieties have been developed that compare favorably with the pure strains. Their great work has been in individual selection and the production of the so-called pedigree barleys. They started by procuring foundation stock from the best varieties in the United States and Canada. These were planted in small plats and watched carefully during their growth. At harvest time the best heads were selected from the best plants. These were in turn grovm in centgener beds and again selected. After five years of this close selection larger quantities of seed were saved and planted. The sixth year small quantities of this seed were sent out to co-operating students all over the state. In this way the new varieties were disseminated throughout the state and were soon available to every farmer in the state. The Wisconsin station bred for large yields, for resistance to smuts, and for uniformity in size and quality of grain, in length and strength of straw and in time of germination and maturity. In twelve years time sixteen different strains of pedigree barley were developed that were superior to the old strains.The six-row bearded was found to be superior to all other varieties. Minnesota has probably done as much or more work than any other State in breeding and improving wheat. The Experiment station has been working for twenty-five years to produce improved strains for the use of Minnesota farmers and have been well repaid in many cases. Their work has been along the usual lines of cross breeding and selection. Two kinds of cross breeding were practiced - out-crosses and in-crosses. Out-crosses were made between plants of different varieties and in-crosses between plants of the same variety. The seed from these crosses was planted and the best one plant from each arose selected to continue the breeding. This individual selection was continued for several years before it was put out as a new variety. In this way several new varieties were developed that yielded very creditably with the best of the old varieties. However, it was found that large numbers had to be worked with in order to get even one cross that was good. The vast majority of them were no good. While out-crosses were harder to get than in-crosses, they gave more evidence of being crossed than did the in-crosses. Only one of the in-crosses showed enough variation from the parent stock to indicate that it was hybrid. Some of the best known varieties originating from crosses are the Fife and Blue stem wheats. After establishing a new variety it was disseminated throughout the state in much the same way as the barley varieties were in Wisconsin. The gain to the farmers of the state by this practice has been immense. It made available for the farmer pedigreed seed that was far superior to the scrub stock commonly used. The Nebraska Experiment station nas done a great deal of wheat breeding work in last twenty years, mainly along lines of selection. In former years the greater part of the wheat grown in Nebraska was spring wheat. The wields were not what they should be so attention was turned to winter wheats. The experiment station took the matter up and has developed a winter wheat that has increased the yield five bushels and more per acre. The variety found to be best suited to Nebraska was the Turkey Red. Dr. T. L. Lyon undertook to improve this variety by a systematic method of selection. To start with he took one thousand select heads and planted them in separate rows in the nursery. The second year the best heads from these rows were planted. This was continued for three years, some of the poorest being discarded each year. At the end of three years. The seed was planted in thirtieth-acre plats. For five years this was continued until good comparisons could be made of the various strains. Then the wheat was sold in ten bushel lots, in different sections of the state on the condition that the buyer should seed according to instructions and make a report on the results. The reports showed that an average of twenty-one tests in twenty counties gave a net increase of four bushels per acre over the local Turkey Red Wheat. In only three instances did the local wheat excel the improved strains. Dr. Hjalmar Wilssen of the Swedish Experiment Station, has done a great deal of breeding work with small grains. He has made selections tending toward improvement along all lines. In some instances he has been successful and in others very unsuccessful. He found that deficiencies of technical nature could be overcome, such as length of spikelet but where the foundation principles were concerned it was more difficult and sometimes impossible. One of his failures was an attempt to breed few stronger culms in barley. One of the most valuable barleys grown in Sweden was disposed to lodge or fall down. Nilsson tried to remedy this by selection. He spent years of hard work on the problem but never accomplished anything and finally gave it up. He later accomplished his end problem by producing another strain of barley which had as good quality as the old, and at the same time had strong culms. / Master of Science