Brass rule against forty-one seedlings

Dear Hooker,

Yesterday I returned to the Ipomoea bed with the same brass rule I used on the 14th, and measured again the forty-one seedlings whose heights I had recorded a week prior, and the exercise has left me uneasy in a way I did not expect. The plants themselves changed, of course, that is the point of the experiment, crossed against self-fertilised, and the crossed run taller as the notes show. What troubles me is the two measurement sessions themselves. Plant 23 in the crossed row I recorded at 42 inches on the first pass; on the second, stretching the rule a shade differently, angling for the true crown of the plant rather than the highest leaf my eye settled on first, I made it 44. Plant 9 in the self-fertilised row moved the other way. Nothing in the plants altered. Something in the measuring did.

I confess I was for a long while inclined to treat these discrepancies as errors to be apologized for, slippage of the thumb, fatigue of the eye, and to be smudged out of the tables. But they accumulate. The seedlings bend toward the light, so the “height” depends on when in the day one measures. The rule’s zero end is worn. My own posture over the bed differs from morning to afternoon. Each alone is trivial. Together they suggest that no single reading of a plant is the plant, and that the honest figure is not one measurement but the drift among several.

It is difficult to avoid the inference that this applies beyond the seedlings. The barnacle dissections had the same character, the same valve yielding slightly different dimensions on second examination, and I recorded then only the more convenient figure. I wonder now how many of my published numbers are single readings wearing the costume of certainty.

So I have decided this: beginning Monday I shall take every height in the bed three times, on three separate occasions, before the figure enters the notebook, and where the readings disagree I shall record the disagreement itself alongside the mean. The spread is not noise to be discarded. The spread may be the most truthful thing the brass rule tells me.