peer reviewed ; This work is on the analysis of automnal data collected during an intensive two-month bird ringing campaign organized near Liège in 1984. At first we remind the reader of some features and results related tho the study of avian migrations in general, and to the chaffinch Fringilla coelebs in particular. This information, gathered from an abundant literature, is essential to the good understanding of our own results. We then deal with three aspects of the data we collected during the fall migration, 1984. The first one describes the visible migration evolution of the chaffinch, as we watched it on the field as it appears for each age and sex class following the evolution curve of the daily number of catches. Furthermore, we compare these results and relate them to the evolution of the meteorological conditions, which have been logged daily on the capture site. Thus, we ascertain that in 1984 the young birds are passing by before the adults and the females before the males (Chi2) and that for a given wind direction the migration intensity is determined by rainfalls. The second aspect is a biometrical principal components analysis of a group of 426 chaffinches. Besides a strongly marked sexual dimorphism in favour of the males this analysis revealed groups, independently from individual age and sex. These groups differ from each other in biometric characteristics and especially in their period of migratory passage. The groups are obviously more or less overlapping. The results of the first aspect are emphasized in the interpredation of the clusters observed on the diagrams. The interpretation of the separating factors is the speculative nature: the wellknow clinal variations and the more or less pronounced migratory habits of the chaffinch as disclosed in the literature are not sufficient to explain the observed trends. Finally, a plot of the daily weighted averages of the wing shape index (Holynsky, 1965), interpreted according to Rensch (1938) confirms beyond any doubt the differences between ...