Notes. The data are ordered with respect to K from 2 to 12 and, within a K ladder, with rising J (see Sect. 1). Column 5: E_low/k: energy above the ground state of the lower level of a given inversion doublet; k is the Boltzmann constant; (E_up – E_low)/k  ~ 1.0–1.5 K. The beam size is θ_b  ~ 50′′  ×  (18 GHz/ν) with ν being the line frequency. Column 6: uncertainties in the flux densities are dominated by calibration errors,  ± 15% (1σ; see Sect. 2) and are thus higher than the formal errors derived from Gaussian fits. Furthermore, if lines are spectrally unresolved, the given peak flux density is only a lower limit. Column 7: from Gaussian fits with formal standard deviations, which are likely underestimating the real errors in case of spectrally unresolved features. Column 8: in many cases, the line width of the maser component is similar or smaller than the channel spacing displayed in Col. 4. In these cases, obtained line widths are upper limits, peak flux densities are lower limits, and no error is given. No maser emission was seen in the (1, 1), (2, 1), (4, 1), (5, 1), (6, 1), (2, 2), (3, 2), (4, 2), (5, 2), (3, 3), (4, 3), (7, 3), (4, 4), (6, 4), (5, 5), (6, 5), (8, 7), (10, 8), (12, 9) and (10, 10) lines down to 3σ limits of order 0.05 Jy in an individual 0.7–1.0 km s^-1 channel.