When Satoshi launched Bitcoin, the currency had virtually no value, and only a few people were mining for new coins. It was just a novelty at that point, with people trading around UTXOs as a way to distribute ownership of the coins and bootstrap interest in the network.

For the first few years of Bitcoin, individual transactions were relatively small from a data standpoint, so the size limit was never an issue since it did not affect the network.

Fast forward a few years, and that bootstrap network started to develop a market price users were exchanging them for fiat currency, which turned into speculating on this asset. Exchanges made it easier for users to get access to trade and transfer Bitcoin, which began to put pressure on the underlying blockchain.

As more people rolled in, transactions started to fill up those 1MB blocks regularly, and something needed to be done to improve throughput.

The obvious option was to increase the block size, which kicked off the block size wars. Satoshi never indicated why the limit was put in place; it is speculated that it was to keep the blockchain small and discourage large amounts of spam transactions.

So, the size was always up for debate; some wanted 4MB blocks, some 8MB blocks, and others wanted a more nuanced approach with Bitcoin’s block weight.

The concept was introduced with the SegWit (Segregated Witness) upgrade in 2017, and it fundamentally changed how Bitcoin measures and limits the size of blocks in the blockchain.

...read more at thebitcoinmanual.com