Focuses on a popular procedure for interpreting principal components (PC). Author's argument that ignoring small-magnitude loading and approximating PC by the linear combination involving only the remaining variables is potentially misleading in various respects; Examples supporting the argument.