The reference number of a formula should remain the same throughout a book.
For example, Gauss' law in integral form is introduced in [Chou, chap.1, p.23]
as (1.27). In [Chou, chap.2, p.72], when Choudhury quotes the same law, he still uses
the same reference number (1.27). This way clearly links the formula to its
origin. In contrast, Wangsness gives a new reference
number every time he quotes the same formula (Compare [Wangs, (10.55)] with [Wangs,
(11.2)]). This is a poor practice, for when we find the formula [Wangs, (11.2)], we
may forget how it is
introduced or derived. Also compare [Jack, (1.44)] with [Jack, (3.126)].
When one can find a reference in a common textbook, one should not refer readers to research journals.
Journals are supposed to be read by advanced graduate students or experts.
Example. It is not difficult to prove the relation given in [Born, p.56, l.-2].
Born should have referred his readers to [Bir, p.37, l.4] instead of a journal.
When we quote a theorem, we should make its references as accessible as
possible. For example, textbooks are more accessible than than journals; recent
publications are more accessible than ancient ones.
Example. It is easier to prove the
formula given in [Born1, p.59, l.-2] using [Bir,