60
RICE UNIVERSITY STUDIES
It should not be thought that the observed long-term exponential
growth of imprints will continue indefinitely. Physical resources,
as well as author resources, cannot maintain the pace of exponen-
tial growth. It follows that the current phase of imprint growth
must ultimately terminate, yielding to a type of growth, or perhaps
decline, that will have an upper bound. It is important to know
whether this time is come, or whether it still is far in the future,
for in the first case we may breathe a sigh of relief, assured that
our resources will come into compass with library requirements
and perhaps even permit the luxury of elaborating access to
library collections in a studied and leisurely manner not subjected
to the current continual strain to catch up with the flood of acces-
sions. In the second case, which the authors believe is the more
likely of the two, exponential growth at least at current (and
historic) rates will continue for a long period, and a relaxation of
current accession pressures cannot be hoped for. In this case, every
effort must be bent to command resources in the most efficient
and rapid manner possible to cope with the information glut. It is
in this connection that the computer must be considered, for the
growth with time of the number of computer operations per sec-
ond and the number of operations per dollar, are both exponential
also and therefore provide the possibility of containing and assimi-
lating the growing flow of information.
Because the current phase of exponential growth must ultimately
change to some type of growth occurring at a lower rate, and
ultimately to a growth or decline which has an absolute upper
bound, it is of some interest to study the likely forms that future
growth curves will take, as well as the way in which one growth
stage will pass into the next. The remaining sections of this paper
are devoted to determining the forms of natural growth curves that
appear pertinent to the problem of library growth so that three
main problems can be investigated. These are :
1. Is the current phase of exponential growth likely to persist
for long?
2. What is the utility of approximating library and related
growth curves by piecewise exponential functions in place of
more complex curves ?
3. What is the nature and interpretation of fluctuations about
the exponential trend ?
The answer to the first problem has the greatest immediate sig-
nificance, for it determines whether future decades have in store