Comment on the AM method