EN
This text is an attempt to apply the concept of a legal text unit to the description of strictly defined structures of legal texts, namely the gerund strings. The main attention is focused on the methods used to identify multi-segment terminological structures and on distinguishing them from strings that constitute a combination of linguistic expressions. The overall conclusion of such analysis is based on the indication of the mechanism used for connecting and limiting gerund structures in this scope. Therefore, this work includes, among others, the formal procedure of verification, which allows for distinguishing, with significant precision, the syntactic constructions with gerund forms and similar sequences of terms.