Cutting Grimm alfalfa in late August and early October prevented the normal fall accumulation of reserve proteins and carbohydrates, and resulted in complete winterkilling of plants during the winter 1934-35.

The roots of the plants which were not cut after Aug. 29 were 75 percent larger than those given a second late cutting on Oct. 8. In addition the roots of the early cut plants contained approximately 30 percent more carbohydrate and nitrogenous reserves on a green weight basis.

The percentage of both soluble and insoluble nitrogen was higher in the uncut plants than in the cut plants throughout the test period. There was no evidence that protein splitting was a factor in cold resistance.

The reducing' sugars constituted a progressively smaller portion of the root reserves as fall advanced. The non-reducing sugars (sucrose) increased during the fall, and were constant during the winter with a slight increase toward spring.

Starch began disappearing in early fall and constituted a very small portion of the root reserves in any variety after October. Digestion was most rapid in the hardy varieties. Analyses for starch in the spring showed no tendency toward reconversion of sugar to starch.

Analytical data on varieties offered no criteria by which the varieties could be placed in their proper order of hardiness as exhibited in the field tests. Non-hardy varieties were consistently higher in some of the reserve fractions than the hardier ones. Hardy varieties were not significantly higher than the non-hardy groups in any of the reserve fractions.

Although possibly important in cold resistance, the pectin and water-soluble gum fractions showed fluctuations among varieties that would not justify the use of these fractions as indexes of the hardiness of varieties at the present time. Attempts to test protoplasm stability directly were unsucessful.

The results of the experiments support the protoplasm differentiation hypothesis of winter hardiness, which assumes that both an available reserve and a genetic ability to use these reserves in the building of a stable protoplasm are essential factors contributing to winter hardiness.