Wholes should equal the sums of which they consist.
		This is the core test by which we construct and focus our measurement systems.    The essential assumption of the test is that a system must be consistent with itself, if it is to be correct. When surveyors or instrument makers perform “loop tests” – testing for the consistency of their measurements along different paths, they are applying this conservation requirement, again and again, finding sources of error, and fixing them.
			When we define finite increment equations for circumstances where several physical effects are at play at once, we need arithmetical procedures that pass loop tests.   (This may not be a sufficient condition for a correct procedure, but it is a necessary condition.