IntroductionOscillatory bio-and chemiluminescence processes, unlike non-oscillatory ones, have not been analysed as yet by means of the memory function approach (MFA) founded on multiplicative stochastic models of photon-counting time series nor have such models been constructed before now. Through the application of the MFA[1,2] to non-stationary non-oscillatory photon emissions from perturbed or stimulated biosystems it is now possible to propose new types of perturbation measures [3,4] rather than relying on more ambiguous or/and local measures [5][6][7][8] which have been employed up until now.The question concerning non-local quantitative characteristics of perturbation/stimulation/excitation arises for biosystems whose internal (quasi-)periodic processes are manifested by oscillatory photon emissions. Periodic processes, both spontaneous and induced by xenobiotics, are quite normal for biosystems, especially since living organisms are open non-linear dynamic systems [9][10][11][12][13][14][15]. From the theoretical viewpoint of dynamic systems, stable or unstable oscillations in biosystems are connected with the existence of such attractors as a centre, limit cycle, or stable/unstable focus, respectively. Equivalently, in the case of oscillatory behaviour one can say there are interactions between subsystems within any given (bio-or eco-) system. That is to say, oscillations in the populations of two competing species[16], oscillations within a single organism [17], oscillations in non-synchronized cell cultures[10], oscillatory immunological processes[10] and oscillatory diseases [18,19].One can therefore expect that a MFA-based comparative analysis of oscillatory and non-oscillatory luminescence processes, resulting from either perturbation or stimulation, can give an additional insight into the problems of perturbation and persistence of biosystems. Such an analysis starting from the calculation of multiplicative stochastic models of photon emission process is presented here. Multiplicative stochastic models[20,21] describe periodic component-containing stochastic processes in the form of a superposition of two autoregressive-integrated moving average processes ARIMA(p,d,q) × ARIMA(P, D ,Q) s , which represent non-periodic and periodic components, respectively.