Turing's original reaction network is systematically studied, particularly
in what concerns: (a) Its ability to produce patterns in a predictable way.
(b) The feasibility of its concentration-independent sink term. Despite th
e widely accepted view that Turing's original model presents some inherent
inability to produce regular structures, the pattern formation properties o
f this model are found to obey the predictions of the corresponding Linear
Stability Analysis in the one-dimension and in 'small' two-dimensional syst
ems. An 'Enzymatic' variation of the original Turing's Model is introduced,
where the unrealistic sink term is substituted by an enzymatic degradation
. It seems that reaction networks of this type can inspire a promising sear
ch for chemical or biochemical experimental systems with pattern formation
properties, even in the absence of high non-linearities. IL is pointed out
that temporal oscillations, impossible for the original Turing's Model, are
stable and persistent in its Enzymatic variation. (C) 2000 Elsevier Scienc
e Ltd. All rights reserved.