The stimulus is talking about lava—specifically, what old, solidified lava can tell us about Earth's magnetic field. This is what we know:
(1) solidified lava is magnetic, aligning with Earth's magnetic field when it solidified;
(2) lava flows with an age difference of millions of years often have different magnetic directions;
(3) lava flows with an age difference of thousands of years often have similar magnetic directions.

From premises (1) and (2), the author draws the sub-conclusion that Earth's magnetic field must change direction over time. And this all leads to the main conclusion that Earth's magnetic field changes direction very slowly.

The argument follows a phenomenon-hypothesis pattern: the author observes a certain phenomenon, and comes to a hypothesis about its cause. Based on phenomena about lava flow magnetism—especially premise (3), that some lava flows thousands of years apart still have similar magnetic directions—the author hypothesizes that Earth's magnetic field's directional shifts must take hundreds of thousands of years to happen.

There are several ways we can weaken a phenomenon-hypothesis argument. We could propose an alternative hypothesis, a competing explanation for the observed phenomenon. For instance, maybe Earth's magnetic field actually shifts really fast, so over a few thousand years it's made a full rotation. We could also introduce contradictory evidence that wouldn't make sense under the author's hypothesis. For example, what if there were also many lava flows only hundreds of years apart with very different magnetic fields?

Because there are several possible weakening strategies, it's important to carefully consider the answer choices and not get too invested in a particular option. That way, we can focus on identifying the actual effect of each answer choice on the argument and avoid making assumptions to justify a trap answer.