Table II. Scores from five samples of black walnuts each cracked by six operators according to scoring schedule II.
| Operators | |||||||||
|---|---|---|---|---|---|---|---|---|---|
| Variety | Location | Year | 1 | 2 | 3 | 4 | 5 | 6 | Average |
| Snyder | Ithaca, N. Y. | '46 | 89.2 | 87.3 | 78.9 | 94.4 | 87.5 | 91.5 | 86.5 |
| Thomas | Ithaca, N. Y. (A) | '46 | 83.5 | 79.2 | 83.1 | 78.0 | 84.2 | 83.8 | 83.6 |
| Thomas | Ithaca, N. Y. (B) | '46 | 73.1 | 67.4 | 73.4 | 74.1 | 69.6 | 83.8 | 73.6 |
| Cresco | Ithaca, N. Y. | '46 | 66.0 | 69.2 | 63.1 | 67.2 | 68.5 | 60.2 | 65.7 |
| Brown | Ohio | '45 | 62.5 | 51.0 | 65.4 | 60.4 | 48.1 | 64.8 | 58.7 |
| Average | 74.9 | 70.8 | 72.8 | 72.8 | 71.6 | 78.8 | 73.6 | ||
| Least significant difference (5%) for variety averages | 6.2 | ||||||||
A third scoring system, involving 1) weight of kernels in grams for the first crack, plus 2) total weight of kernels, 3) all divided by the number of marketable pieces (as counted following sifting on a ¼" round hole screen) was tried, and the resulting ranking of the varieties was very similar to that obtained with systems I and II. The results from this system appeared to be the most precise, but it was not considered as generally acceptable as system II, since the latter would be easier to record and calculate.
It is the opinion of the authors that Schedule II gives a score that estimates very well the relative merit of the samples tested as to crackability, yield and marketability. It is simple to use and the only equipment required is a scale accurate to 1/10 gram. Calculations are reduced to a minimum and the characters used are not dependent on judgment of the individual making the test. It should be pointed out, however, that differences in score of less than six points are not significant on the basis of testing done to date. As more tests are made this value may be reduced. The schedule should serve as a measure to establish differences between varieties, particularly when a considerable number of tests are made. It can also be relied upon to measure differences due to the location of trees of the same variety, variation of the same variety from year to year in the same and in different locations and differences of a similar nature. In ranking varieties which have scores within the limits of variability, it will be necessary to use judgment as to small differences of appearance. No scoring schedule can be expected to entirely eliminate the judgment of experts. Also it must be realized that characters other than the nuts, such as bearing habit, hardiness, yield of trees, disease resistance and the like must be considered in finally establishing the value of a variety.