What does train_data.batch(BATCH_SIZE) create?

Good morning. I thought that the dimension of the batch for the training data had to be said inside the training part of the code. If I have for example a train_data of 1000 elements and I use train_data.batch(200), does this return a list of 5 lists containing the training data splitted? And why don't we specify the dimension of the batch inside the training part of the code, instead? Thank you