On the usefulness of finding charts Or the runaway carbon stars of the Blanco & McCarthy field 37

We have been recently faced with the problem of cross–identifying stars recorded in historical catalogues with those extracted from recent fully digitized surveys (such as DENIS and 2MASS). Positions mentioned in the old catalogues are frequently of poor precision, but are generally accompanied by finding charts where the interesting objects are flagged. Those finding charts are sometimes our only link with the accumulated knowledge of past literature. While checking the identification of some of these objects in several catalogues, we had the surprise to discover a number of discrepancies in recent works.The main reason for these discrepancies was generally the blind application of the smallest difference in position as the criterion to identify sources from one historical catalogue to those in more recent surveys. In this paper we give examples of such misidentifications, and show how we were able to find and correct them.We present modern procedures to discover and solve cross–identification problems, such as loading digitized images of the sky through the Aladin service at CDS, and overlaying entries from historical catalogues and modern surveys. We conclude that the use of good finding charts still remains the ultimate (though time–consuming) tool to ascertain cross–identifications in difficult cases.