“…On the other hand, it is our experience that the typical magnitudes of p 0 and dY are of the order of 10 À4 [11,13,14] and so, as a first approximation, we can estimate kðTÞ with the accuracy better than 10%, when simply neglecting the term p 0 =2. Under similar circumstances, the same approximation has been often used within the HAUP technique [26,27]. Using the kðTÞ data obtained this way, the standard results for the linear birefringence Dn and the relation g 11 = n n ¼ 2k Dn, it is easy to arrive at the temperature dependence of the circular birefringence g 11 = n n shown in Fig.…”