On the disappearance from the blood of intravenously injected insulin