The urgent problem of the philosophy of science is to ascertain whether the scope of new knowledge in natural sciences may happen to be restricted a priori, or before any experience. And if so, what are the considerations delimiting the scope of any possible knowledge? The paper deals with this important problem promoting one helpful analogy: it is argued that the relation of a priori knowledge to new knowledge must be represented as relation of metalanguage to object language. This decision helps the author to demonstrate how the scope of new knowledge in natural sciences is actually restricted by a priori knowledge of nature, or the concept of Nature, in Kantian sense. In addition, the author tries to define some risks of the science of nature associated with radically new forms of knowledge being accepted.