My Twitter feed today highlighted an item about the use of card indexes in the Library of Congress, and it brought back many memories from my early career in the information profession. In 1970 I started my career as an Information Officer (which sounded very important) at the British Non-Ferrous Metals Research Association in 1970. The BNF (as it was usually referred to) appointed Robert Hutton as Director in 1921. As well as being a distinguished metallurgist he was also very committed to technical information and libraries. A photograph of the Library does not give an adequate indication of its scale and the quality of the library team. The information team of two were early students of Jason Farradane and provided me with an excellent initial training. The information office was on the ground floor (see the previous link) at the end of the electrochemistry laboratories, the smell of which pervaded the entire floor.

Card indexes were everywhere but the most precious was the Trade Name Index. This had been started in around 1925 and by the time I arrived had perhaps 5000 “5 by 3” cards. On each of these cards was not only what we knew about a trade name but annotations (always with initials) by people who had used the card and had more information to add. The index was open to anyone in the BNF. Even now creating a computer database would be a huge challenge because of the very wide range of content on the cards. Some might have official standards numbers, some telephone numbers and some hand-drawn phase diagrams of alloys. No card was ever removed; it was not uncommon to find that I was looking at a card that had last been updated ten years’ earlier – the information was still current and useful in an industry several hundred years old. The first book on mining and metallurgy appeared in 1556!

We only had one failure. In the early 1950s we were asked about a source of Red Spot Copper and we were still looking for it when I left in 1974. We were aware of red rot, which is when brass corrodes and the original red shines through, but not a brand of copper. We followed up many leads, all neatly noted on the card, which was actually three cards clipped together as members of staff added a fresh lead.

The cards not only contained information about the trade name but about who had asked us about it. On one memorable occasion a rather difficult BNF member asked us about a trade name and one of my colleagues laconically asked him if this was the same name he had asked us about the previous year and two years earlier. Although they had limitations (notably going missing through being refiled in the wrong place) card indexes were a very useful tool for information work. I hate to think how long it would take to design a database and also to provide a very usable interface for additions and corrections.

I should add for the record that the BNF also used punched feature cards, both edge-punched (developed by E.G.Brisch in the early 1950s) and optical coincidence cards, a commercial version of the cards designed by the Polish cryptanalysts who completed the first decryption of German Enigma traffic in 1936.

Martin White