As some readers may have observed already, the Niels Bohr Archive (NBA) has made a good start at making information about its collections, and even some of the collections themselves, available on the internet. The work is still in progress, but just for this reason it may be useful at this point to share some of the lessons learned and to receive feedback from our users in order to make improvements.

The story begins in the spring of 2003, when the Danish State accepted the NBA’s application for a pilot project. The purpose was, first, to make the in-house catalogues of our archival collections — which existed in various formats, both on paper and as computer files — available to everyone on the internet from our website. Second, we wished to make some of the archival material itself digitally available for researchers upon application. For our paper collection we chose the Niels Bohr Political Papers, which is of a manageable size, which had just been released by the family for research use, and which we deemed of special interest. In addition, we applied to make a good part of our film and sound collections similarly available, both because some of this material was in a deteriorating physical state and because we wanted to try out several kinds of collections.

The scanning of the documents could begin at once as an in-house project, which at the same time made it possible to continue paying salaries to our permanent staff. The scanning had two purposes. First, we made high-quality digital copies in TIFF format to be kept as a security backup of the collection itself. From these, we then made black-and-white PDF-files of significantly lower resolution with the intention of making them available to researchers as digital documents. The sound recordings and films were digitized out of house. Just as in the case of the documents, two sets of files were produced: one of high quality for backup purposes, and one of lower quality for later use by researchers.

But how best to make the material available to researchers in practice? This was a puzzle that troubled us for a long time, while our small staff was also occupied with other matters, such as completing the Niels Bohr Collected Works. In early 2007 my old friend Joe Anderson at the AIP Center for History of Physics alerted us to the existence of the archives software Archon, freely available from the University of Illinois. However, it took another year and a half just to be able to install the software, a process which proved to require extensive assistance from the infrastructure of our host institution, the Niels Bohr Institute (NBI). Nor was learning how to find one’s way around in the software, once installed, a trivial matter. In the end we succeeded, with crucial help from the Archon developers themselves.

However, the exercise proved to be worth it. Entering the information of our archival material on the collection level, which was already on our homepage, into standard archives fields in Archon proved a trivial matter. But our more detailed catalogues—previously only for staff use and maintained in a variety of formats sometimes down to the document level—proved a real challenge. With the help of computer experts at the NBI, I was able to reeducate myself in programming, and after several months the majority of our already existing catalogues was imported into Archon. As a consequence, a user can now easily search and move around in and across catalogues.

Fortunately, Archon allows linkage to digital documents outside the software itself, and we were thus able to link from the document level of the catalogues to each archival document. Unlike the catalogues, however, the archival documents are generally restricted. Any person visiting our archives database is therefore required to apply for access to such documents. This is done by downloading, printing and submitting an application from our website, much like the application that researchers visiting the NBA in person need to fill out. Upon approval of the application, the researcher will receive a password giving access to archival material for a limited time for the particular project described in the application.

The pilot project was finally completed early this year, and we are grateful to the Danish Ministry of Science for generously allowing us to reach this stage in spite of the many delays along the way. Yet, the project must still be considered to be in its beginning stages. For one thing, our current version of Archon on the internet is not the most recent one; updating to the latest version, which we expect to occur within half a year’s time, will lead to significant improvement, not least with regard to speed of access.

Also, we will continue to add collections and make catalogues more detailed and precise. Most importantly, perhaps, the only archival documents available so far are the ones digitized as part of the pilot project. We are currently scanning the letters in the extensive (close to a thousand correspondents) Bohr Private Correspondence, which documents Bohr’s large network outside physics, both inside and outside Denmark, and hope to make this material available in a year.

Our flagship, the Niels Bohr Scientific Correspondence (BSC — about 400 correspondents), is part of the Archives for the History of Quantum Physics, which is available on microfilm at several institutions around the world. Digitizing this collection is therefore seen as less urgent. However, over the years many new scientific letters have come to light, so that there now exists a supplement to the BSC, with as many correspondents as the original collection, which has not been microfilmed. The NBA has recently received a grant from the Ministry to digitize this supplement, the documents of which we hope to make available on our website by 2013, the centennial for Bohr's model of the nuclear atom.

We have found the task of digitizing our collections generally complex, sometimes frustrating, and in the end satisfying. At this stage it is particularly important to receive feedback from our users about any aspect of our Archon website, archon.nbi.dk (which can also be reached through our normal web address, nba.nbi.dk). Does Archon (which while we worked with it received the prestigious Mellon Award for Technology Collaboration in 2008) seem to be an appropriate means to make our collections available? Are the collections easy enough to navigate? Which collections need to be digitized next? Is the reproduction quality (usually in black-and-white) of the documents themselves acceptable? We would be grateful for any comment to these and other questions sent to our email address nba@nbi.dk. Obviously, nothing can compete with a research visit to the NBA, which we still encourage!