Przibram's Rule slates that the ratio of the mean weights in successiv
e instars of hemimetabolous insects is 2.09. Faber (1994) described a
two-stage approach to testing Przibram's Rule when the instars of the
measured individuals are unknown. In the first stage, individuals are
assigned to one of the instars on the basis of their weights. In the s
econd stage, a lest of the null hypothesis that the means of logarithm
ic weight in the two groups differ by the logarithm of 2.09 is perform
ed. This approach suffers from two problems. First, errors in assignin
g individuals to instars will exaggerate the difference between instar
s. Secondly, the validity of the test in the second stage is based on
the implicit assumption that the variance of logarithmic weight in the
two groups is equal. This note describes a test of Przibram's Rule th
at avoids these problems.