The short answer
Read an immunology experiment — a titre curve, a knockout panel, a transfer — and work out which part of the response is running and which component is missing.
Written and checked by GAMSAT tutors — not AI-generated.
Try the reasoning style
We treat forgetting as a failure — a lapse to be patched with reminders and records. Yet a mind that kept everything could not think; it would drown in the undifferentiated noise of every moment it had ever lived. To forget is not so much to lose information as to decide, mostly without our noticing, what was never worth keeping.
The author's argument relies most directly on which unstated assumption?
Pick an option to see how the tutor reasons to the answer — not just whether you were right.
Not quite — the answer is B.
Work backwards from the conclusion: a mind that ‘kept everything’ supposedly ‘could not think.’ That only follows if thinking means leaving most of experience out — so B is the premise the argument quietly rests on. A raises reliability, which the passage never weighs; C contradicts ‘mostly without our noticing’; D smuggles in a claim about intellect the passage never makes. The question rewards finding the hidden premise, not recalling a fact.
Immunology arrives as an experiment you have never seen. The question is never what an antibody is, but which part of the response ran and which is missing. Antigen never instructs a cell; specificity is shuffled beforehand, so it selects those that already fit.
- anti-X
- anti-Y
How to read an immunology result
Read the lag
Hours means innate: fast, fixed, identical every exposure. About a week means adaptive from scratch. Two to four days means the clones were already there.
Then the height
Titres sit on log axes, so read multiples, not distance: the second anti-X peak looks modestly taller and is sixty times the first, 100 to 6000.
Demand a control antigen
Faster-and-bigger is memory only if it is specific. Anti-Y entered the same animal on day 28 and still took a week to reach 100, so the acceleration is X's alone.
Ask what transfers
Serum adds antibody, not cells; it carries complement too, but the recipient already has its own. Serum protecting a naive recipient shows antibody is sufficient; only cells restoring it means the break is cellular.
A flat column localises nothing
"No antibody detected" is an output, not a location: it fits no B cells, no T-cell help, or no class II. Find the second thing measured, a graft, a transfer, a second antigen, because that column discriminates.
| Mouse line | Anti-X titre, day 14 | Anti-X titre, day 35 | Skin graft |
|---|---|---|---|
| Wild type | 100 | 6000 | rejected, day 11 |
| P — no mature B cells | not detected | not detected | rejected, day 11 |
| Q — no mature T cells | not detected | not detected | intact at day 100 |
| R — no MHC class II | 15 | 20 | rejected, day 16 |
Worked example: what serum carries
A mouse recovers from bacterium Z. Its serum goes into a naive mouse, which survives a Z challenge. Ten weeks later that recipient is challenged again and dies. Why?
Check yourself
A guinea pig is injected with protein M on day 0 and again on day 40; on day 40 it also receives protein N, which it has never met. Anti-M peaks at 150 on day 14, falls to 45 by day 40, then peaks at 4200 on day 45. Anti-N is first detectable on day 43 and peaks at 3800 on day 47. Which conclusion do these data support?
Key takeaways
- Antigen selects clones that already exist; it never instructs them.
- Read the lag before the height; on a log axis, read height as multiples.
- Faster and bigger is memory only if a fresh antigen in the same animal refuses the boost.
- One readout cannot localise a break: ask what else was measured.
Practise this with real GAMSAT-style questions
Free account: a timed diagnostic, an AI tutor that explains every answer, essay marking on the official rubric, and a plan built around your weak spots.