The short answer
Removing a proposed cause before observing an effect can be more informative than collecting another striking image. The inference depends on what the intervention removes and what it leaves unchanged.
Evidence and original analysis
Make the intervention assumption visible
Invented counts with unchanged uncharged nuclei.
| Removal model | Implied charged count | Consistent with total 20? |
|---|---|---|
| 100% clearing | 14 | Yes |
| 70% clearing | 20 | Yes, at the boundary |
| 50% clearing | 28 | No |
Chronology
A timeline of the turning points
Early work reported
Ionisation and condensation investigated.
Retrospective lecture
A selected account of the clearing control.
The physical contrast
Wilson reported that an electric field applied before expansion removed ion-produced condensation nuclei, so their cloud did not form. Uncharged nuclei could also be produced, but were not removed by the field.
The intervention tested a property of the nuclei. It was different from deciding the direction or charge of a particle from an already recorded curved track. The later chamber applications should not obscure this earlier experimental question.
Sources for this section: [1]
An invented matched comparison
Consider two perfectly matched samples. Without clearing, 20 droplets are counted; after clearing, six. Stipulate one counted droplet per nucleus, no change to uncharged nuclei and complete removal of charged nuclei. Infer the charged and uncharged counts.
Now weaken just one assumption: suppose the field removes only half the charged nuclei. Can the same counts still be explained without changing another assumption? Find the minimum removal efficiency compatible with these observations.
Sources for this section: [1]
Reference answer
Complete removal gives 14 charged and six uncharged nuclei: 70% of the initial count is charged under this model. This is a constructed count decomposition, not a calibration of Wilsonโs apparatus.
Let efficiency e be the removed fraction of C charged nuclei. The observed reduction gives eC=14. Since C cannot exceed 20, e must be at least 14/20=0.7. At e=0.5, C would be 28, exceeding the original total. That version of the model is inconsistent.
Sources for this section: [1]
What the model cannot supply
The calculation is deterministic and assumes perfect matching. Real counting efficiency, droplet merging, incomplete clearing or changed condensation conditions could break its interpretation. A cloud image without these controls is not automatically a census of ions.
Sources for this section: [1]
Continue the chronology
Related history guides
Biography context
Explore United Kingdom through connected lives
Connected biographies
The researcher and the record
Quick answers
Common questions
- Is 70% a historical experimental result?
- No. It follows only from the invented counts and specified model.
- Does clearing test the same question as track curvature?
- No. This task changes nuclei before condensation rather than interpreting an existing track.
Primary and institutional sources
How this history was checked
Selected historical passages are distinguished from the original teaching model. Invented inputs are not reconstructed experimental data. This article was reviewed on 31 October 2026.