A photograph is a memory of a surface. A notebook is a memory of what you thought the surface was, at a time when you can still be shown wrong. Both are useful. Only one of them records the question.

What goes on the page

Date and time, written out. Place, specific enough that a stranger could stand there: the path, the side of the tree, the pot on the third bench, not “campus.” Weather and recent disturbance if they might matter — rain last night, a mowing, a window left open. A sketch with a scale, even a bad one. A sketch forces you to see whether the leaf is opposite or alternate. A photograph lets you postpone seeing.

Then the names you are willing to defend, and the names you are not. “Crow-sized black bird” is a better datum than a guessed species written without a question mark. Uncertainty belongs in the ink. Correcting it later in a different pen is a record of learning, not a shame.

The question, written first

Before the count, write what you are comparing and what result would change your mind. “I think the north side of the wall has more lichen because it stays wet. If the south side has as much or more, I am wrong, or wetness is not the reason.” You can now come home with a result you do not like, which is the only kind of result that teaches. The biologist’s corner says the same thing for a laboratory project. The notebook is where you prove you meant it.

Afterward

Copy nothing into a clean version that silently fixes the awkward lines. If you transcribe for a report, say that you transcribed, and keep the original. Number the pages. Do not tear one out because a stranger might read it. A notebook you would be willing to show a supervisor, blanks and wrong turns included, is the beginning of a trustworthy method. The rest of this site — the experiments especially — is that habit, practiced by people who had better flasks and the same problem.