The board in T 925/98 noted that, according to the respondent, the range 30% to 50% given in claim 1 infringed Art. 123(2) EPC 1973, since such a range was not disclosed in the originally filed documents of the patent in suit, which only disclosed a general range of 30% to 60% and a preferred range of 35% to 50%. The board held, however, that, according to the established case law, in the case of such a disclosure of both a general and a preferred range, a combination of the preferred disclosed narrower range and one of the part-ranges lying within the disclosed overall range on either side of the narrower range was unequivocally derivable from the original disclosure of the patent in suit and thus supported by it (see T 2/81, OJ 1982, 394; T 201/83, OJ 1984, 481; and T 53/82, T 571/89, T 656/92, T 522/96 and T 947/96 all referring to T 2/81). Thus claiming a range from 30-50% did not contravene Art. 123(2) EPC 1973. In the case in point, moreover, graphs indicated that the claimed range was in fact the most efficient one. This frequently cited decision was referred to, for example, in T 328/10, T 2001/10, T 227/13; see also T 1107/06.

In T 249/12 the board held that the conclusion of T 2/81 according to which the amendment made was allowable was arrived at by inter alia considering that the new range was "unequivocally and immediately apparent to the person skilled in the art"; according to the board in T 249/12 this was equivalent to the "gold standard".

In T 1170/02 the board held that the combination of the lower limit of the general range with the lower limit of the preferred range, thus excluding the preferred range, did not fall under the principles developed in decision T 2/81 (OJ 1982, 394). In such a case, it might be useful to ask whether the skilled reader of the parent application would seriously contemplate working in the range referred in the divisional application or alternatively whether there was anything in the parent application as filed or his common general knowledge which would cause him to exclude the possibility of working in that range (with reference to T 187/91, OJ 1994, 572). To the board, the skilled person would, in view of the data in the parent application, seriously consider working beyond the lower limit of the preferred range as well. Under these circumstances, the range could be directly and unambiguously derived from, and is consistent with, the parent application as originally filed. See also T 1389/08 and T 205/13.

In T 612/09, however, the board observed that T 2/81 had established a two-step approach. In the first step, it had considered that the two part-ranges of the general range (in T 612/09: 3 to 75 mg/kg) lying outside the preferred range (in T 612/09: 10 to 25 mg/kg) would be unequivocally and immediately apparent to the person skilled in the art. It had then considered that no new matter was introduced by combining the preferred range with the upper part-range. The board in T 612/09 considered that only the first step of the analysis carried out in decision T 2/81 was necessary to arrive directly and unambiguously at the claimed range. Indeed, the board held that the two part-ranges lying within the overall range on either side of the narrower range (a dose of 3 to 10 mg/kg and a dose of 25 to 75 mg/kg) were directly and unambiguously disclosed to the person skilled in the art. See also T 2159/11.

In T 1919/11 the subject-matter of claim 1 of the main request related to silver at a concentration of at least 1µM to less than 200µM. In the description, it was stated in two separate consecutive sentences that "When silver is incorporated in the medium, it will be added at a concentration of less than 900 µM, preferably less than 500 µM, and more preferably less than 200 µM" and "When silver is incorporated in the medium, it will be added at a concentration of at least 10 nM, preferably 100 nM, more preferably 1 µM, and typically at 10 µM". The board observed that the situation in the case at issue was not comparable with that in T 1107/06 and dealt with in the settled jurisprudence of the boards of appeal. A general range, which means a lower limit which is unequivocally combined with an upper limit, and a preferred disclosed narrower range – equally consisting of a lower limit which is unequivocally combined with an upper limit – were simply missing. Even a kind of parallel structure in indicating the upper and lower limits (less/at least, preferred or more preferred) implied no unequivocal correlation between a particular upper limit and a particular lower limit because there was no teaching that such an arrangement was intended. Therefore, one of the upper limits mentioned in the first sentence in the description of the parent application as originally filed (as cited above) and one of the lower limits mentioned in the second sentence were arbitrarily combined, which did not represent a direct and unambiguous disclosure.

In T 1990/10 the board had to decide whether the application as filed provided a basis for the temperature range "below 35°C" in claim 1. The application as filed disclosed various temperatures, both specific temperature values ("30°C") and temperature ranges, such as open-ended ranges ("below 37°C") and closed ranges with defined upper and lower-end values ("30°C to 35°C"). The board considered that the term "below" was explicitly disclosed only for defining the broadest mentioned temperature range, namely "below 37°C". The board held that "below 35°C" was also not implicitly derivable from the broadest open-ended range "below 37°C" in combination with the upper-end value of the closed range "30°C to 35°C". Applying the criteria of T 2/81, the combination of the lower-end and the upper-end values of the closed range with the broadest temperature range would result in the temperature ranges "30°C to below 37°C" and "35°C to below 37°C", not however, in the open-ended temperature range "below 35°C". In addition, the closed temperature range "30°C to 35°C" included the specific temperature "35°C" whereas the open-ended range "below 35°C" in claim 1 explicitly excluded this value. It was also not a combination of ranges and sub-ranges as explained in T 2/81 to merely transfer the term "below" in "below 37°C" to the upper end value of "30°C to 35°C". In the absence of any indication in the application as filed to do so, this transfer of the term "below" to another temperature value or temperature range had no basis in the application as filed. The board also had to decide whether the application as filed provided a basis for the temperature range "about 26°C to 32°C", which it rejected.

In T 249/12 the board allowed the amendment to the range of "10-50 mass %", which combined the use of the upper limit of the less preferred broader range (50 mass %) and the lower limit of the most preferred narrower range (10 mass %). Since both end-points of the new range were both specifically mentioned in the application as filed, the new range was "unequivocally and immediately apparent to the skilled person" (T 2/81), i.e. the range was directly and unambiguously disclosed in the application as filed. See also T 1143/17 which relates to the combination of the lower end-point of a preferred range with the upper end-point of a more preferred range.

In T 1320/13 claim 1 defined a specific ratio in the range of "between 0.6 and 1". The appellant argued that the range had a basis in the disclosure of the range "0.1 and about 1" in claim 5 in conjunction with the value "0.6" disclosed in a list of individual values, which read: "about 1.0, 0.9, 0.8, 0.7, 0.6, 0.5, 0.4, 0.3, 0.2, 0.1 or even 0, derivable therein". The board held that T 2/81 did not apply in the case at issue. The skilled person would not regard the list of individually disclosed values of ratios on page 5 of the application as individual end-points of ranges. Firstly, a list of individual values – even if disclosed as here in descending order – did not relate to values that lie between them, while a range necessarily encompasses all the values that lie between its two disclosed end-points, i.e. a list of individual values is conceptually different from a range. Secondly, the list contained no pointers to a particular combination of ratio values. Accordingly, a specific selection of values also did not clearly and unambiguously emerge for the skilled person from the content of the application.

In T 703/16 the board held that a peptide fragment constituted a discrete physical entity, made up of individual amino acids and having defined physical end-points, which was not comparable to a range of a quantitative parameter within which the invention can be performed (distinguishing its case from T 2/81 and T 1063/96).