Quick Navigation


Case Law of the Boards of Appeal

2.1.8 Error margins and definitions of limits

Prior to the opinion given in G 2/98 (OJ 2001, 413), it was sometimes possible to claim error margins or definitions of limits which differed in the subsequent application from the original one. In the jurisprudence which came after G 2/98, however, reference is no longer made to the decisions handed down in T 212/88 (OJ 1992, 28), T 957/91, T 65/92 and T 131/92 (see "Case Law of the Boards of Appeal of the EPO", 3rd ed. 1998, p. 263 et seq.). But since the facts are only comparable up to a certain point, it is not possible to say unequivocally to what extent the principles laid down earlier still apply.

In T 201/99 the cited priority applications disclosed, in respect of the mean residence time of the material in the granulator/densifier, either the range "from about 1-6 minutes" (claim 1 of both priorities) or, for all the examples, a specific residence time of "about 3 minutes". In contrast, claim 1 of the granted patent claimed a mean residence time of "1-10 minutes". The appellant had argued that the upper limit of "6 minutes" was not disclosed in the cited priorities as a cut-off value (i.e. as possibly related to the function of the invention and its effect) and thus constituted a feature which might be modified without changing the nature of the invention. The board, however, held that, even if for the sake of argument the range of "about 1 to 6 minutes" given in the priority applications was interpreted as an open range extending far beyond "6 minutes", the value of "10 minutes" as well as, for instance, the range from above 6 minutes up to 10 minutes still constituted selections not disclosed in the priority applications. It followed that for a skilled person applying his common general knowledge the value of "10 minutes", at least, was not derivable directly and unambiguously from the prior applications.

The invention in T 423/01 concerned a method for analysing length polymorphisms in simple or cryptically simple DNA regions. According to claim 1, the DNA sequences in the invention incorporated repeating units of between three and six nucleotides. The respondents objected that the claimed upper limit, in particular, could not be found in the original application documents. The board considered T 201/83 (OJ 1984, 481) to be relevant to the question of rightful priority claims, even though it concerned the original disclosure of quantitative ranges of values within the meaning of Art. 123(2) EPC 1973. According to this decision, an amendment to a range in a claim is allowable on the basis of a particular value described in a specific example, provided the skilled man could have readily recognised this value as not so closely associated with the other features of the example as to determine the effect of that embodiment of the invention as a whole in a unique manner and to a significant degree. In the case in hand, the board had already established in the context of Art. 123(2) EPC 1973 that the phrase "approximately 6 to 10" explicitly disclosed the number six. A range between the value 3, which was disclosed as preferred in Example 2 (identical to Example 2 of the priority document), and the lowest value of the "range" already mentioned in the priority document for the upper limit value (at least one nucleotide and a maximum of 6 to 10 nucleotides) was to be considered disclosed in the original application. The claimed priority was valid.

In T 250/02 the board found that the subject-matter of a claim for a herbal essential oil in which the total amount of carvacrol and thymol was at least 55%, and preferably 70%, by weight of the said essential oil could not be derived directly and unambiguously from a priority document disclosing a herbal oil containing "thymol and carvacrol in levels of 55% to 65%".

Further decisions concerned with differing ranges indicated in a priority document and disputed sets of claims include T 903/97, T 909/97, T 13/00, T 136/01, T 788/01, T 494/03 and T 537/03. See also T 110/07, reported in point 2.1.9 below.