Is the ratio between number of citations and publications cited a true constant

The first experimental citation indexes of scientific literature were compiled almost twenty years ago. By the time we had completed the 1964 Sczknce Citu-tion Inde# (SCF'), we were aware that there was a surprising near-constancy in the ratio of 1.7 between references processed each year and the number of different items cited by those references. Very early we began to call the 1.7 ratio the citation constam. We have frequently discussed this number at lSI@ , and used it for various estimates informally, I have never attempted any rigorous analysis of it, though it has been mentioned in some of these essays. If you examine any annual SC1 Guide, this 'constant' is readily apparent in the chronological statistical analYsis provided. As the number and type of journals covered by K'] has grown, the ratio has changed slightly. Perhaps my own mathematical and statistical naivct~ has made it possible for me to suffer in silence so many years while I wondered about the probability of a true constant. This does not mean that a number of people have not concerned themselves with regularities in citation data. Probably Derek Price was the first in recent times to publish on the subject, though he himself often cites the pioneer studies of A .J. Lotka in this connection. 1 Using SCZ data, Price showed in 1965 how many papers will be cited n times. 2 Authors often ask me how sigrtiflcant it is that a paper has been cited ten Number 6 times in one year. They are surprised to learn that less than 25% of all papers will be cited ten times in all eternity! How categorical can you get ? As you have seen in the various lists of highly cited papers we have published in C#ment Content@ (CG'), any paper cited ten times in one year is i~~o fato significant. Occasionally there is an anomaly. But a paper cited ten times in each of two successive yearn is well on its way to citation stardom. Whether the author is on the way to immortality depends on how well he or she does in other papers. But let us return to Gaty$e//s 'con-stant'. Why is it 1.7? A Dutch scientist, M.C. Gomperts, studied the constancy of citation in the special field of Chladni's plates,3 but he failed to take note of our magical 1.7. A.E. Cawkell reviewed Gompert's paper in some detail …