Indications on bearing habits were obtained for a six-year period on 96 trees, Nos. 1 through 140 (Table 1). Crop records for each of these trees were examined for relatively high and low yield by seasons. Convincing evidence on the alternation of bearing has accumulated during this six-year period with 46 percent of the trees having lighter crops every other year. Of these, 28 trees bore lighter crops in the odd years and 16 trees bore lighter crops in the even years. Tree 117, previously mentioned as outstanding in regard to yield, produced lighter crops in 1940, 1942, and 1944. This tree is located in west Tennessee. Walnut trees bearing lighter crops in 1941, 1943, and 1945 are more abundant in the eastern than in the western portion of the Tennessee Valley. This occurrence undoubtedly accounts for much of the variation found between the eastern and western portions. Four other yield patterns were recognized in 30 per cent of the trees. These indicate the existence of uniform annual crops and three-year cyclic bearing of black walnut. The bearing habits of the remaining 24 per cent of the trees is considered merely irregular, since definite patterns cannot be recognized until bearing records cover a longer period of years. Nut Quality The cracking quality of the nuts from the trees in this study was tested on a random sample of nuts from each crop that was collected and brought back to Norris. The nuts of each sample were weighed and the average nut weight computed. The nuts were then cracked in a hand-cracking machine, and kernels that could be extracted with the fingers were removed and weighed.[13] From this weight was computed the first-crack marketable kernel percentage. The nuts that still contained kernels were recracked and the remaining kernel removed. All kernels, including crumbs, were then weighed in order to compute the total kernel weight and kernel percentage. Finally, all of the quarters extracted were counted, and the average number of quarters was computed. Kernels recovered at first crack and the average number of quarters extracted indicate the relative ease of extraction of kernels. Cracking quality of walnuts for individual sample trees averaged by crop years are presented in Table 1. Nuts of all crops collected from four trees, 57, 58, 60, and 139, were shriveled or abnormal, and afforded no test of nut quality during the six-year period. Thus, nut quality data, based on 440 nut crop samples, are complete for 128 of the 132 sample trees. From this study, the average common black walnut in the Tennessee Valley has a nut weight of 17 grams, a kernel weight of 3.3 grams, a total kernel content of 20 per cent, a marketable kernel recovery at first crack of 17 per cent, and a quarter recovery of unbroken quarters averaging 1.8. [Footnote 13: The kernels were extracted over a 6-mesh wire screen. In commercial cracking, kernel pieces passing through this type of screen are not marketable as kernels.]