These cross-directional interactions will generate the incessant fluctuations
 in the biocoenosis that creates a perception of a form of balance, because our
 macroscopic impression is that the association is stable, without noticeable
 changes. Therefore, a complete defoliation, due to its highly visible
 consequences, creates an impression that something out of the ordinary
 happened and, to explain this, it seems logical to assume a disturbance of a
 balance.
 
The gradating population, though, does not drop out of the biocoenosis,
 and continues to be under the influence of its factors (Glen, 1954). The effect
 of biotic factors (e.g. parasites and predators) is greatest at high densities
 (“density dependent mortality factors’, Smith 1935). The high activity of
 obstant elements during gradations of Lymantria or Aporia is very well
 known; several authors see this as a self-regulating ability of the biocoenosis,
 to maintain some sort of balance (Friederichs, 1930; Schwenke, 1953).
 
To this we respond as follows: the richer the biocoenosis, the more
 complicated are the interactions among its members, and more mortality
 factors are likely to decimate every member population. This can explain the
 inverse relationship between density fluctuations and the complexity of the
 biocoenosis (Solomon, 1949). This is also supported by the observation that
 there are more frequent gradations in agrobiocoenoses than in coenoses
 under lower human influence (Schimitschek, 1942). Is it correct to conclude
 from this that there is a lack of biological equilibrium in the former?
 
At first sight, it is attractive to explain the gradations of corrumpents in
 agrobiocoenoses through disturbances of the equilibrium (Friederichs, 1930).
 The writer himself interpreted insect pest damage this way (Szeényi, 1944),
 and he may still retain this opinion if: 1) he could define, precisely, the
 equilibrium, and; 2) his later studies, carried out in biocoenoses less influenced
 by humans, had not convinced him that gradations of corrumpent elements
 are no less frequent in such communities. Point 1 cannot be satisfied, because
 the studies mentioned under point 2 do not support the steady state of the
 biocoenosis, or any kind of balance. We illustrate this with two examples.
 One of them are the results for a gall fly, Janetia (Arnoldia) cerris, studied
 over 10 years in the same area that also exhibited gradations in this system,
 and where fly abundance can reach a level whereby their galls cause growth
 disorders on the host plant. Two years after the latest gradation (1955), the
 abundance became so low that only one Janetia gall was found per several
 hundred leaves. The second example is the Rubus-Crataegus-Rosa bushland
 on the southern slope of the Nagyszénäs, above Nagykovacsi, that was
 censused over several years, and where, in 1953, without any previous signs,
 the abundance of Cydia tenebrosana (Laspeyresia roseticolana) reached
 extremely high levels, but hardly any were found in the subsequent year.
 
These biocoenoses are not studied by anyone, while the agrobiocoenoses
 are always under observation, and as their corrumpents are mostly
 economically damaging animals and their activity creates attention. We may