There are two different ways of defining complexity. 1) Traditionally, the word "complexity" is considered synonymous to "organization". The transformation of species is an expression of victory against random indifferencism. 2) The means of measuring complexity that was conceived by Kolmogorov has the advantage of having an external reference. Therefore, its logical reliability is considerable. But Kolmogorov's complexity will be at its best in cases of pure randomness. These mutually incompatible definitions explicitly demand a classification system. The first definition of complexity is contrary to the second one. This must be explained more precisely in order not to disturb the logic of Kolmogorov's conception and to enable to add to this conception, as closely as possible, along the logical rules derived from Godel's incompleteness. The author proposes a beginning of complexity typology, founded on the necessity of laws that rule a random substract aiming at organization. The generality of laws to be selected will have a direct effect on the logical strength of the "biologists' definition" of complexity. Two fundamental laws are expressed, one derived from mathematics and the other from physics, two fields alien to biology. This fact improves the logical accordance to the argument.