|
| 18 x 38 x 33000 / 90000 = 257.4. Valid if there are 90,000 kernels to the bushel. Invalid if there are smaller kernels and/or the sample was biased. We tend to skip over the small or missing ears and so inflate this guess. Some would use 80K kernels per bushel and some would use 105K. The sample would be a little more valid if you shelled all your sample ears, dried them to 15% (perhaps in the microwave but some may pop) and weighed the total, computed an average per ear, multiplied by the population and divided by 56 pounds to the bushel. This would take out most of the error of seed size but won't cure the sample bias.
Most times this formula is quite a bit optimistic.
Gerald J. | |
|