We studied the citation histories of 326 papers published in 1961 and cited in the 18 subsequent years. On the average the citations reached a maximum five years after publication, followed by a slow decline such that 20 years after publication the citation rate is half the maximum rate. Theoretical papers show the same rate of decline as observational ones. Theoretical and observational papers are cited with equal frequencies. The most frequently-cited papers are almost invariably long ones, although only half of the long papers are cited frequently. Frequently-cited papers have even a longer lifetime, namely about 27 years to decline to half the maximum citation rate. Various additional statistics and caveats to these conclusions are given.