We present an analysis of the behaviour of solar activity and look for
the presence of low-dimensional deterministic chaos within it. The or
iginal databases for such analysis have been the daily sunspot number
(1818-1990) and daily sunspot areas(1874-1989) from which we have cons
tructed twenty different data sets, raw and filtered, displaying the s
olar and magnetic cycle. We have used the Grassberger-Procaccia algori
thm to compute the correlation dimension which, also, has allowed us t
o obtain the K-2 entropy and, for some time series, the maximum Lyapun
ov exponent has also been computed. Our results show that in none of t
he twenty time series considered does evidence appear of chaotic behav
iour, since there is no saturation of the correlation dimension with t
he embedding dimension and the K-2 entropy shows a divergent behaviour
. A study of previous works which claim this kind of behaviour to be p
resent in solar activity suggests that such a conclusion has been deri
ved from very short scaling regions obtained using low time delays in
the computations of the correlation dimension. The behaviour of solar
cycle, with periods of low activity, suggests the presence of determin
istic chaos and some of its features can be reproduced by means of sim
ple nonlinear dynamo models. However, it seems that for an unambiguous
detection of such behaviour, from solar activity data or proxy record
s, we will have to wait for the availability of longer and reliable da
ta sets covering the periods of reduced activity.