The State of São Paulo imports straw of Sorghum for its broom industry when this material could be satisfactorily produced in the State. The principal hereditary defects of the broom-corn which could be eliminated to a certain extent by selection are goose neck panicles, twisted panicles and stemmy panicles. Various introduced varieties were subjected to comparative field tests along with a local variety which came years ago from Argentine. This variety superseeded all the others when subjected to selection. All the lineages derived from it when tested with the primitive and recently introduced stock revealed far superior in many respects. Seeds of these lineages were turned over to the competent Department of the Instituto Agronómico to be increased in guantity and then distributed to farmers. In this work the X² test is recommend for analysis and interpretation of the results.

The leaf area per stalk of five varieties of grain sorghum differing in maturity, height, and size and number of leaves were measured during five consecutive years. The total leaf area per acre ranged from about 21,000 to more than 190,000 square feet. The average leaf area per stalk ranged from about 0.64 to 5.742 square feet, depending on the variety and season. The approximate requirement for the production of 1 bushel of grain was 4,000 to 11,400 leaves, having a total area of 570 to 3,800 square feet. Thus, for each leaf per stalk functioning during the fruiting period the grain yield was 3.5 bushels to 6 bushels per acre. Abundant rainfall during the vegetative period stimulated leaf development. Less leaf area was required to produce a bushel of grain in a dry year than in a wet year, but the highest yields were obtained in seasons of abundant rainfall.

1. Five sets of 100 seeds each of 11 species of weeds were used in these tests each year for a period of 7 years. 2. The seeds were placed in the silage in different silos and at different locations in the silos where they remained for periods ranging from 33 to 1,636 days. 3. Five of the 11 species used show some germination after storage in the silo. Field bindweed germinated in 6 of the 7 years, while annual morning glory and velvet leaf germinated in 5 of the years. 4. Field bindweed gave an average germination of 16% after passing through the silo compared with 27% when stored outside exposed to the weather. Velvet leaf gave higher germination after storage in the silo than when stored outside. 5. Twenty-eight seeds comprising three different species germinated after approximately 4 1/2 years in the silo. 6. The viability of rough pigweed, yellow foxtail, Johnson grass, smartweed, sunflower, and cocklebur was apparently destroyed in the silo. The seed of field bindweed, velvet leaf, morning glory, giant ragweed, and barnyard grass, however, seemed able to resist the effect of ensiling. 7. The danger of spreading certain species of weeds by means of corn and sorghum silage crops is reduced but not eliminated by ensiling the crop.