I am seeking clarification regarding the statement:

"Since records can span two blocks and we want to avoid missing any records that span blocks, we will consider that each read might include only $2$ records to ensure we don't miss records spanning blocks."

Given that each block is $1100$ bytes and each record is $350$ bytes, three records $($totaling $1050$ bytes$)$ fit entirely within a single block, leaving $50$ bytes unused. Therefore, records do not span across blocks. Could you please explain why the consideration is to include only $2$ records per read instead of $3?$