Some Cautionary Notes on Making Inferences about Relative Brain Size