When memories did?

when memories did? did what?

Context:

Glanzman's team found the same mechanism held true when studying the snail's neurons in a Petri dish. The researchers placed the sensory and motor neurons that mediate the snail's withdrawal reflex in a Petri dish, where the neurons re-formed the synaptic connections that existed when the neurons were inside the snail's body. When serotonin was added to the dish, new synaptic connections formed between the sensory and motor neurons. But if the addition of serotonin was immediately followed by the addition of a substance that inhibits protein synthesis, the new synaptic growth was blocked; long-term memory could not be formed.
The researchers also wanted to understand whether synapses disappeared when memories did. To find out, they counted the number of synapses in the dish and then, 24 hours later, added a protein synthesis inhibitor. One day later, they re-counted the synapses.
What they found was that new synapses had grown and the synaptic connections between the neurons had been strengthened; late treatment with the protein synthesis inhibitor did not disrupt the long-term memory. The phenomenon is extremely similar to what happens in the snail's nervous system during this type of simple learning, Glanzman said.